1972_GE_SCR_Manual_5ed 1972 GE SCR Manual 5ed
User Manual: 1972_GE_SCR_Manual_5ed
Open the PDF directly: View PDF 
.
Page Count: 712
| Download | |
| Open PDF In Browser | View PDF |
SCR MANUAL
FIFTH EDITION
GENERAL _
ELECTRIC
Prepared by Application Engineering Centers
Auburn, New York
and
Geneva, Switzerland
Editors:
D. R. Grafham
Mgr., Application Eng.-Geneva
J. C. Hey
Mgr., Application Eng.-Auburn
Contributing Authors:
A. P. Connolly
R. W. Fox
F. B. Golden
D. R. Gorss
S. R. Korn
R. E. Locher
S.J.Wu
Layout Design:
D. K. Barney
Production:
S. Babiarz
D. Farrell
D. G. Seefeld
SEMICONDUCTOR PRODUCTS DEPARTMENT
GE-NERAL
ELECTRIC
ELECTRONICS PARK, SYRACUSE,N. Y.13201
SCR MANUAL
The circuit diagrams included in this manual are for illustration of typical semiconductor applications and are not intended as constructual information. Although reasonable care has been taken in their preparation to assure
their technical correctness, no responsibility is assumed by the General Electric Company for any consequences of their use.
The semiconductor devices and arrangements disclosed herein may be
covered by patents of General Electric Company or others. Neither the disclosure of any information herein nor the sale of semiconductor devices by
General Electric Company conveys any license under patent claims covering
combinations of semiconductor devices with other devices or elements. In the
absence of an express written agreement to the contrary, General Electric
Company assumes no liability for patent infringement arising out of any use
of the semiconductor devices with other devices or elements by any purchaser of semiconductor devices or others.
Copyright © 1972
by the
General Electric Company, U.S.A.
Electronics Park
Syracuse, N.Y. 13201
!
II
FOREWORD
TO THE FIFTEENTH ANNIVERSARY EDITION OF THE SCR MANUAL
Publication of this 5th Edition marks fifteen years since General
Electric introduced the first commercial SCR. Though still a teenager,
the SCR has grown to be the most prominent semiconductor device for
static power conversion and control.
The fast-growing success of the SCR is paralleled by the growth
of the General Electric SCR Manual. First published as an application
note in 1958, the General Electric SCR Manual has been periodically
revised, maintaining the basic theme of a practical, rather than theoretical, circuit and application guide for design engineers, students,
teachers, and experimenters. This Edition is written by a group of
engineers who, figuratively, live in the solid state power conversion
arena. They are in constant touch with equipment designers and, as
such, are exposed to the varied circuit problems and decisions peculiar
to the application of power semiconductor devices. These authors have
gained their insight and experience by contributing to literally thousands of successful design projects involving thyristors in addition to
drawing on the experience and work of their predecessors.
The previous Edition has been completely reviewed in detail.
Much that is new has been added, reflecting the polish that SCR applications have acquired in the past five years. These changes do not stand
out. as new chapters, rather, each chapter has had the additional or
revised information blended in with that which remains current and
valid. For those of you who have used previous editions the format
remains intact to maintain any familiarity you may have developed.
The dual trends of increasing performance in military and industrial SCR's on the one hand and the shift to fabrication techniques
which lend themselves to very high volume production of consumer and
light industrial SCR's on the other hand are evident in the revisions to
this manual. Considerably more detail is included on the parameters of
SCR's along with application tips for the high performance SCR's.
Information showing the designer how to approach optimum utilization
of high volume, plastic encapsulated SCR's is also provided. Still, overall, considerable effort has been spent in keeping the SCR Manual concise and general in nature. For those desiring in-depth treatment of
highly specialized subjects, we refer them to the comprehensive application notes listed on page ??
I sincerely hope that you will find this new Manual useful and
informative.
.
H. D.Culley
General Manager
Semiconductor Products Department
Syracuse, New York
III
SCR MANUAL
This 5th Edition of the SCR Manual is dedicated to one of the
most competent and diverse groups of engineers ever listed on one
page . . . the previous contributors to the SCR Manual, who helped
lay the foundation for this fascinating technology.
A.A. Adem
J. L. Brookmire
J. H. Galloway
F. W. Gutzwiller
E. K. Howell
D. V.Jones
H. Kaufman
H.R. Lowry
N.W.Mapham
J. E. Mungenast
R. M. Muth
T. A. Penkalski
G. E. Snyder
T. P. Sylvan
E. E. Von Zastrow
IV
Introduction
INTRODUCTION
We have not changed the SCR Manual for the sake of change.
Rather we have folded into the fourth edition answers to questions
which you, the equipment designers have been asking for the past few
years. We have certainly included the device and circuit innovations of
these past years. Finally, we have continued our past policy of presenting information in as clear, concise and uncomplicated a fashion as
possible.
HOW TO LEARN ABOUT THE SCR
If you, the reader, are already familiar with the SCR and wish
guidance in the design of practical applications, this Manual is ideal
for your needs. If you wish more detailed information on a specialized
subject, consider the references listed at the end of each chapter, as
well as the comprehensive list of General Electric application notes
(p.658) which are available on request.
If you wish to explore thyristors in a more analytical sense, either
as a semiconductor or as a circuit element, we refer you to "Semiconductor Controlled Rectifiers ... Principles and Applications of p-n-p-n
Devices," a book published by Prentice-Hall, Englewood Cliffs, New
Jersey.
If you have heard of the SCR but would like to start from scratch
in learning how it can help you, we suggest that you obtain a copy of
the General Electric "Electronics Experimenters' Circuit Manual." This
manual describes some 40 ingenious circuits and projects useful in
v
SCR MANUAL
teaching the fundamentals of electronics while constructing projects
having lasting value in the automobile, home, workshop and campsite.
This book was' written by our application engineers on the assumption
that the reader, although learned in his own field of competence, is
new to the SCR and other semiconductors as well.
If you are really impatient to see how a basic thyristor circuit can
work with your equipment load, or if you have no facilities to assemble
your own circuit from electronic components, you may wish to experiment with one of several standard assemblies available from your G-E
distributor. A typical standard triac variable voltage control is shown
below. They are described further in Chapter 7.
A BRIEF DESCRIPTION OF THE SCR
The SCR is senior and most influential member of the thyristor
family of semiconductor components. Younger members of the thyristor family share the latching (regenerative) characteristics of the SCR.
They include the triac, bidirectional diode switch, the silicon controlled
switch (SCS), the silicon unilateral and bilateral switches (SUS, SBS),
and light activated devices like the LASCR and LASCS. Most recent
additions to the thyristor family are the complementary SCR, the
programmable unijunction transistor (PUT) and the "assymmetrical
trigger."
Let's go back to the head of the family after whom this Manual
was named. The SCR is a semiconductor . . . a rectifier . . .a static
latching switch . . . capable of operating in microseconds . . . and a
sensitive amplifier. It isn't an overgrown transistor, since it has far
greater power capabilities, both voltage and current, under both continuous and surge conditions, and can control far more watts per dollar.
VI
Introduction
As a silicon semiconductor-the SCR is compact, static, capable of
being hermetically sealed, or passivated, silent in operation and free
from the effects of vibration and shock. A properly designed and fabricated SCR has no inherent failure mechanism. When properly chosen
and protected, it should have virtually limitless operating life even in
harsh atmospheres. Thus countless billions of operations can be
expected, even in explosive and corrosive environments.
As a rectifier-the SCR will conduct current in only one direction.
But this serves as an advantage when the load requires DC, for here
the SCR serves both to control and rectify-as in a regulated battery
charger.
As a latching switch-the SCR is an ON-OFF switch, unlike the
vacuum tube and transistor which are basically variable resistances
(even though they too can be used as on-off switches). The SCR can
be turned on by a momentary application of control current to the gate
(a pulse as short as.a fraction of a microsecond will do), while tubes or
transistors (and the basic relay) require a continuous ON signal. In
short the SCR latches into conduction, providing an inherent memory
useful for many functions. The SCR can be tutned ON in about one
microsecond, and OFF in 10 to 20 microseconds; further improvements
in switching speed are being made all along.
Just as a switch or relay contact is commonly rated in terms of
the current it can safely carry and interrupt, as well as the voltage at
which it is capable of operating, the SCR is rated in terms of peak
voltage and forward current. General Electric offers a complete family
of SCR's with current carrying capacities from % amp to 1600 amps
RMS, and up to 2600 volts at this writing. Higher voltage and current
loads are readily handled by series and parallel connection of SCR's.
As an amplifier-the smallest General Electric SCR's can be
latched into conduction with control signals of only a few microwatts
and a few microseconds duration. These SCR's are capable of switching
100's of watts. The resulting control power gain of over 10 million
makes the small SCR one of the most sensitive control devices available. With a low cost unijunction transistor firing circuit driving the
larger SCR's, stable turn-on control power gains of many billions are
completely practical. This extraordinary control gain makes possible
inexpensive control circuits using very low level signals, such as produced by thermistors, cadmium sulfide· light sensitive resistors, and
other transducers.
Most of the foregoing list of assets of the SCR apply equally well
to the other members of the thyristor family as you will see in this
Manual. Meanwhile the shortcomings and limitations of thyristors
become less significant as the years pass. Newly introduced high voltage
and bidirectional types lift the transient and operating voltage barriers.
High speed thyristors allow operation at ultrasonic frequencies and
under severe dynamic conditions, and lower semiconductor costs perniit use of higher current rated thyristors instead of critically designed
and expensive overcurrent protection systems.
VII
SCR MANUAL
Best of all, SCR's and thyristors for every type of application ...
industrial, military, aerospace, commercial, consumer ... are more
economical than ever. Best indication of this is the rapidly increasing
tempo of applications of the new plastic-encapsulated thyristors in high
volume consumer applications where every penny is critical.
Here are just some of the conventional types of controls and elements that thyristors are busy replacing and improving upon:
Thyratrons
Relays
Magnetic Amplifiers
Ignitrons
M-G Sets
Rheostats
Power Transistors
Motor Starters
Transformers
Limit Switches
Constant Voltage Transformers
Saturable Reactors
Contactors
Variable Autotransformers
Fuses
Timers
Vacuum Tubes
Thermostats
Mechanical Speed Changers
Centrifugal Switches
Ignition Points
Welcome to the exciting world of the thyristor family, its circuits
and applications. Please bear in mind that the material in this Manual
is intended only as a general guide to circuit approaches. It is not allencompassing. However, 'our years of experience in offering application help have shown that, given some basic starting points like those
in this Manual, you .the circuit designer inevitably come up with the
best, and often unique, approach for your particular problems.
We, in particular, would like to direct your attention to Chapter
20, "Selecting the Proper Thyristor and Checking the Completed Circuit Design." Here, we have tried to pull together a roadmap of the
route to successful design with thyristors avoiding the pitfalls we, and
others, have learned the hard way.
You will find the "Application Index" on pages671a useful guide
in starting the search for the ideal circuit for your application. The list
of Application Notes starting on page674 will also provide specialized
help beyond the detail of this Manual.
VIII
TABLE OF CONTENTS
TABLE OF CONTENTS
1. CONSTRUCTION AND BASIC THEORY OF OPERATION .
. .......... .
1.1
What is a Thyristor? .................... .
1.2
Classification of Thyristors ..................... .
1.3
Two Transistor Analogy of PNPN Operation ....... .
1.4
Reverse Blocking Thyristor (SCR) Turn-Off
Mechanism
1.5
Improvements for Dynamic SCR Operation ..
V-I Characteristics of Reverse Blocking Triode or
1.6
................ .
Tetrode Thyristors
Gate Turn-Off Switch or Gate Controlled Switch.
1.7
Thyristor Used as a Remote Base Transistor.
1.8
Thyristor Construction . . . . . . . .
. ........ .
1.9
Pellet Fabrication ............... .
1.9.1
Pellet Encapsulation . . .
..... ...
1.9.2
1.10
Comparison of Thyristors With Other Power
Semiconductors
1
1
1
1
4
4
10
12
13
14
14
16
20
2. SYMBOLS AND TERMINOLOGY ....... .
2.1
Semiconductor Graphical Symbols.
2.2
SCR Terminology ........... .
2.2.1
Subscripts .................... .
2.2.2
Characteristics and Ratings ..
2.2.3
Letter Symbol Table '"
2.2.4
General Letter Symbols ...
23
23
27
27
27
33
34
3. RATINGS AND CHARACTERISTICS OF THYRISTORS.
35
35
36
37
37
37
38
3.1
3.2
3.3
3.4
3.4.1
3.4.2
3.4.3
3.5
3.5.1
3.5.2
3.5.3
3.5.4
3.5.5
3.6
3.6.1
3.6.2
3.6.3
3.7
3.7.1
3.7.2
Junction Temperature ........ .
....... .
Power Dissipation
Thermal Resistance ... . ..... .
Transient Thermal Impedance ....... .
............ .
Introduction
The Transient Thermal Impedance Curve ..
The Effect of Heatsink Design on the Transient
Resistance Curve
........
Recurrent and Non-Recurrent Current Ratings .....
Introduction
......... .
Average Current Rating (Recurrent).
RMS Current (Recurrent) ....
Arbitrary Current Waveshapes and Overloads
(Recurrent) ..... .
Surge and I 2t Ratings (Non-Recurrent).
Basic Load Current Rating Equations ............ .
................. .
Introduction
Treatment of Irregularly Shaped Power Pulses Approximate Method ................... .
Resistance Welding Ratings for Recurrent
......... .
Pulse Bursts
Recurrent and Non-Recurrent di/dt Ratings ....... .
............ .
Introduction
Industry Standard dildt Rating (Recurrent) ... .
41
42
42
43
44
45
45
47
47
47
51
52
52
53
IX
SCR MANUAL
3.7.3
3.7.4
Concurrent dildt Rating (Recurrent). . . . . . . . ..
Industry Standard di/dt Rating (Gate
Triggered - Non-Recurrent) ..............
Industry Standard di/dt Rating (V(RO)
Triggered - Non-Recurrent) ..............
Turn-On Voltage .........................
High Frequency Current Ratings. . . . . . . . . . . . . . ..
High Frequency Sinusoidal Waveshape
Current Ratings ........................
High Frequency Rectangular Waveshape
Current Ratings ...... ;.. . . . . . . . . . . . . . ..
Voltage Ratings ..............................
ReverseVoltage (VnnM ) and (VnsM). . . . . . . . . ..
Peak Off-State Blocking Voltage (V DRM). . . . . ..
Peak Positive Anode Voltage (PFV). . . . . . . . . ..
Voltage Ratings for High Frequency,
Blocking Power Limited SCR's. . . . . . . . . . ..
Rate of Rise of Off-State Voltage (dv/dt) ......... ,
Static dv/dt Capability. . . . . . . . . . . . . . . . . . . ..
Reapplied dv I dt .........................
Triac Commutating dv I dt. . . . . . . . . . . . . . . . ..
Gate Circuit Ratings. . . . . . . . . . . . . . . . . . . . . . . . . ..
Holding and Latching Current ..................
Reverse Recovery Characteristics. . . . . . . . . . . . . . . ..
54
4. GATE TRIGGER CHARACTERISTICS, RATINGS AND METHODS . .......... .
71
71
73
73
3.7.5
3.7.6
3.8
3.8.1
3.8.2
3.9
3.9.1
3.9.2
3.9.3
3.9.4
3.10
3.10.1
3.10.2
3.10.3
3.11
3.12
3.13
4.1
4.2
4.2.1
4.2.2
4.2.3
4.3
4.3.1
4.3.2
4.3.3
4.3.4
4.3.5
4.3.6
4.4
4.5
4.6
4.7
4.8
4.9
4.10
4.11
4.12
4.13
4.13.1
4.13.2
x
The Triggering Process. . . . . . . . . . . . . .. . . . . . . . ..
SCR Gate-Cathode Characteristics.. . . . . . . . . . . . . .
Characteristics Prior to Triggering. . . . . . . . . . . .
Characteristics at Triggering Point. . . . . . . . . . . .
Characteristics After Triggering. . . . . . . . . . . . ..
Effects of Gate-Cathode Impedance and Bias. . . . . . .
Gate-Cathode Resistance .................. .
Gate-Cathode Capacitance ................ .
. Gate-Cathode Inductance ................. .
Gate-Cathode LC Resonant Circuit .......... .
Positive Gate Bias ........................ .
Negative Gate Bias ....................... .
Effects of Anode Circuit Upon Gate Circuit ....... .
DC Gate Triggering Specifications. . . . . . . . . . . . . . . .
Load Lines ................................. .
Positive Gate Voltage That Will Not Trigger SCR .. .
Pulse Triggering ............................ .
Anode Turn-On Interval Characteristics. . . . . . . . . . .
Simple Resistor and RC Trigger Circuits .......... .
Triggering SCR With a Negative Pulse ........... .
AC Thyratron-Type Phase Shift Trigger Circuits ... .
Saturable Reactor Trigger Circuits .............. .
Continuously Variable Control .............. .
On-Off Magnetic Trigger Circuits ........... .
55
55
55
56
56
58
60
61
61
61
62
63
64
65
66
67
67
68
74
74
76
76
77
78
79
79
81
84
85
85
87
87
90
91
94
95
95
96
97
TABLE OF CONTENTS
4.14
4.l4.1
4.l4.2
4.14.2.1
4.14.2.2
4.14.3
4.14.3.1
4.14.4
4.14.5
4.14.6
4.14.7
4.14.8
4.14.9
4.15
4.15.1
4.16
4.17
4.18
4.18.1
4.18.2
4.19
Semiconductor Trigger-Pulse Generators ....... .
Basic Relaxation Oscillation Criteria. . . . . . . . . .
Unijunction Transistor ................. .
Basic UJT Pulse Trigger Circuit .. _
Designing the Unijunction Transistor
Trigger Circuit . . . . . . . . . . . . . . . .
Programmable Unijunction Transistor (PUT) ...
Designing the PUT Relaxation Oscillator
and Timer Circuits ...
Silicon Unilateral Switch (SUS) ............. .
Silicon Bilateral Switch (SBS) ... _.... _
Bilteral Trigger Diode (Diac) ............... .
Asymmetrical AC Trigger Switch (ST4) ....... .
Other Trigger Devices. . . . . . . . _
Summary of Semiconductor Trigger Devices ...
Neon Glow Lamps as Trigger Devices ....... .
Neon Lamp Trigger Circuits ........... .
Pulse Transformers ......._
Synchronization Methods
Trigger Circuits for Inverters. _
Transistorized Flip-Flops ................. .
PUT Flip-Flop Trigger Circuit ... .
Pulse Amplification and Shaping .. ..
5. DYNAMIC CHARACTERISTICS OF sca's ................... .
5.1
SCR Tum-Off Time, t q . . . . . . . . . . . . . . . .
5.1.1
SCR Tum-Off Time Definitions .. .
5.1.2
Typical Variation of Tum-Off Time .. .
5.1.3
Circuit Tum-Off Time (t,.) .......... .
Feedback Diode ........................ .
5.1.4
5.2
Tum-Off Methods ............... .
Current Interruption ..................... .
5.2.1
5.2.2
Forced Commutation .............. .
5.3
Classification of Forced Commutation Methods ..
5.3.1
Class A - Self Commutated by Resonating
the Load .......................... .
5.3.2
Class B - Self Commutated by an LC Circuit.
5.3.3
Class C - C or LC Switched by Another
Load-Carrying SCR .................... .
5.3.4
Class D - LC or C Switched by an
Auxiliary SCR .................... .
5.3.5
Class E - External Pulse Source for
Commutation _..... _.......... .
5.3.6
Class F - AC Line Commutated.
5.4
Rate of Rise of Forward Voltage, dv/dt.
5.4.1
Reapplied dv I dt _..............
5.5
Rate of Rise of On-State Current, dil dt ..
5.5.1
Solutions to the dil dt Problem .............. .
5.6
Reverse Recovery Characteristics. . . . . .
5.7
Capacitors for Commutation Circuits _..
98
98
100
102
103
10.5
106
109
110
110
111
112
112
113
114
115
117
118
118
119
119
123
123
124
125
127
127
127
128
128
128
128
129
131
132
134
138
139
140
141
141
142
143
XI
SCR MANUAl
6. SERIES AND PARALLEL OPERATION ...........................
6.1
Series Operation of SCR·s ......................
6.1.1
Need for Equalizing Network. . . . . . . . . . . . . ..
6.1.2
Equalizing Network Design. . . . . . . . . . . . . . . ..
Static Equalizing Network. . . . . . . . . . . ...
6.1.2.1
6.1.2.2
Dynamic Equalizing Network. . . . . . . . . ..
6.1.2.3
Other Voltage Equalizing Arrangements. ..
6.1.3
Triggering Series Operated SCR·s. . . . . . . . . . ..
6.1.3.1
Simultaneous Triggering Via Pulse
Transformer .......................
Simultaneous Triggering hy Means of Light
6.1.3.2
Slave Triggering for Series SCR·s. . . . . . . ..
6.1.3.3
6.1.3.4
The Triggering Pulse .................
6.2
Parallel Operation of SCR's ....................
6.2.1
SCR Transient Turn-On Behavior. . . . . . . . . . ..
6.2.2
Direct Paralleli~g Using SCR's With Unmatched
Forward Characteristics and No Sharing
Networks .............................
6.2.3
Use of SCR's With Matched Forward
Characteristics .........................
6.2.4
External Forced Current Sharing. . . . . . . . . . . ..
Triggering of Parallel ConnectedSCR's. . . . . . . . . ..
6.3
149
149
150
152
152
155
159
160
160
161
163
165
165
166
168
173
176
178
7. THE TRIAC ...........................................
7.1
Description ..................................
7.1.1
Main Terminal Characteristics. . . . . . . . . . . . . ..
7.1.2
Gate Triggering Characteristics ..............
7.1.3
Simplified Triac Theory ........... : ........
7.1.4
Commutation of Triacs ....................
7.1.5
Triac Thermal Resistances ..................
7.2
Use of the Triac .............................
7.2.1
Static Switching ..........................
7.2.2
Firing With a Trigger Diode ................
7.2.3
Other Triggering Methods. . . . . . . . . . . . . . . . ..
7.3
Triac Circuitry ...............................
181
181
182
183
184
186
188
189
189
191
192
192
8. STATIC SWITCHING CIRCUITS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
8.1
Introduction .................................
8.2
Static AC Switches ...........................
8.2.1
Simple Triac Circuit and Inverse-Parallel
("Back-to-Back") SCR Connection. . . . . . . . ..
8.2.2
Static Switching With Separate Trigger Source.
8.2.3
Alternate Connections for Full Wave AC
Static Switching . . . . . . . . . . . . . . . . . . . . . . ..
8.2.4
Triac Latching Technique. . . . . . . . . . . . . . . . ..
8.2.5
AC Static SPDT Switch. . . . . . . . . . . . . . . . . . ..
8.3
Negative Half Cycle SCR Slaving Techniques. . . . ..
8.3.1
SCR Slaving and Zero Voltage Switching ......
8.4
"One-Shot" SCR Trigger Circuit. . . . . . . . . . . . . . . ..
8.5
Battery Charging Regulator. . . . . . . . . . . . . . . . . . . ..
8.6
DC Static Switch.. . . . . . . . . . . . . . . . . . . . . . . . . . ..
8.6.1
DC Latching Relay or Power Flip-Flop ..... "
195
195
195
XII
195
197
197
199
200
200
201
202
203
204
205
TABLE OF CONTENTS
8.7
8.7.1
8.7.2
8.7.3
8.7.4
8.7.5
8.8
8.8.1
8.8.2
8.8.3
8.9
8.9.1
8.9.2
8.10
8.10.1
8.10.2
8.10.3
8.10.4
8.10.4.1
8.10.4.2
8.10.5
8.10.6
8.11
8.12
8.12.1
8.12.2
8.12.3
8.12.4
8.12.5
8.12.6
8.13
8.14
8.14.1
8.14.2
Flasher Circuits ..............................
DC Flasher With Adjustable On & Off Time. ..
Low Voltage Flasher. . . . . . . . . . . . . . . . . . . . . ..
Sequential Flasher .'. . . . . . . . . . . . . . . . . . . . ..
Low Power Flasher. . . . . . . . . . . . . . . . . . . . . . ..
AC Flasher ..............................
Protective SCR Circuits. . . . . . . . . . . . . . . . . . . . . . ..
Overvoltage Protection on AC Circuits. . . . . . ..
SCR Current-Limiting Circuit Breakers. . . . ..
High Speed Switch or "Electronic Crowbar". ..
Ring Counters ...............................
Cathode Coupled Ring Counter. . . . . . . . . . . . ..
Anode Coupled Ring Counter ...............
Time Delay Circuits .........................
UJT ISCR Time Delay Relay. . . . . . . . . . . . . . ..
AC Powered Time Delay Relay. . . . . . . . . . . . ..
Ultra-Precise Long Time Delay Relay. . . . . . ..
Time Delay Circuits Utilizing the Programmable
Unijunction Transistor (PUT) ..............
30 Second Timer. . . . . . . . . . . . . . . . . . . . ..
Long Delay Timer Using PUT ..........
A 60-Second Time Delay Circuit Switching AC.
One Second Delay Static Tum-Off Switch. . . .
N anoampere Sensing Circuit With 100 Megohm
Input Impedance. . . . . . . . . . . . . . . . . . .
Miscellaneous Switching Circuits Using Low
Current SCR's .............................
Dual Output, Over-Under Temperature Monitor
Mercury Thermostat/SCR Heater Control. . . ..
Touch Switch or Proximity Detector. . . . . .
Voltage Sensing Circuit ...................
Single Source Emergency Lighting System.
Liquid Level Control .....................
Thyratron Replacement . .
. . . . . . . . . . . . . . . . ..
Switching Circuits Using the C5 or C106 SCR as a
Remote-Base Transistor. . . . . . . . . . . .
"Nixie"@ and Neon Tube Driver. . . . . . . . . .
Electroluminescent Panel Driver. . . . . . . . .
9. AC PHASE CONTROL ... .................................
9.1
Principles of Phase Control ....................
9.2
Analysis of Phase Control ......................
9.2.1
AC Inductive Load Phase Control. . . . . . . . . . ..
9.2.2
Using Thyristors on Incandescent Lamp Loads.
9.3
Commutation in AC Circuits. . . . . . . . . . . . . . . . . . ..
9.4
Basic Trigger Circuits for Phase Control. . . . . . . . .
904.1
Half-Wave Phase Control ..................
9.4.2
Full Wave Phase Control ..................
9.5
Higher "Gain" Trigger Circuits for Phase Control. ..
9.5.1
Manual Control .................... . . . . ..
9.5.2
Ramp and Pedestal Control
. . . . . . . . . ..
@ Trademark Burroughs Corporation
205
205
206
207
208
208
209
209
210
212
213
213
214
215
215
216
217
218
218
219
219
220
221
222
222
223
224
224
225
226
227
228
228
228
231
231
232
241
245
246
249
249
252
254
254
256
XIII
SCR MANUAL
9.5.3
9.5.4
9.5.5
9.6
9.7
9.7.1
9.7.2
9.7.3
9.8
9.8.1
9.9
9.9.1
9.9.2
9.9.3
A Wide Range Line - Voltage Compensation
Control ...............................
3 KW Phase Controlled Voltage Regulator ... "
860 Watt Limited-Range Low Cost Precision
Light Control .. . . . . . . . . ; . . . . . . . . . . . . . ..
Trigger Circuits for Inductive AC Loads ...........
Phase Control With Integrated Circuits ...........
The PA436 Monolith Integrated Phase-Control
Trigger Circuit . . . . " . . . . . .. . . . . . . . . . . ..
Circuit Design With the PA436 ............ "
PA436 in High Power Circuits ..............
Typical Phase-Controlled Circuits for DC Loads. . ..
A 1.2 KW; 60 V Regulated DC Power Supply ..
Polyphase SCR Circuits ...................... "
Simple Three Phase Firing Circuit (25% to
100% Control) .........................
Full Range Three Phase Control System. . . . . ..
Use of the PA436 Phase Control Integrated
Circuit in Three Phase Circuits . . . . . . .
10. MOTOR CONTROL EMPLOYING PHASE CONTROL. . . . . . . . . . . . . . . . . ..
10.1
Introduction .................................
10.2
Brush-Type Motors Controlled by Back EMF
Feedback ..........................
Half-Wave Universal Series Motor Controls.
10.2.1
10.2.2
Full-Wave Universal Series Motor Control. ....
10.2.3
Shunt Wound and PM Field Motor Control. . ..
10.3
Brush-Type Motor Control- No Feedback. . . . . . ..
10.3.1
Half-Wave Drive for Universal, Shunt or
PM Motors .' . . . . . . . . . . . . . . . . . . . . . . . . ..
10.3.2
Full-Wave AC Drive for Universal Series
Motor ..............................
10.3.3
Full-Wave DC Motor Drives ............
10.3.4
Balanced-Bridge Reversing Servo Drive.
lOA
Induction Motor Controls ..............
10.4.1
Non-FeedbackControIs .................. '.
1004.2
Indirect Feedback ...... .................
1004.3
Spe()d Regulating Control of Induction Motors .
10.5
Some Other Motor Control Possibilities ...........
10.5.1
Single-Phase Induction Motor Starters .......
11. ZERO VOLTAGE SWITCHING .. .. .. ................
11.1
Introduction ....... ..................
11.2
Electromagnetic Interference ...................
11.3
Discrete Zero Voltage Switching Circuits .....
11.3.1
Basic Switching Circuit ....................
11.3.2
Two Transistor Switching Circuit .. ...
11.3.3
CSCR Zero Voltage Switch. . . . . . . . . . . . . . . ..
11.3.4
Triac Zero Voltage Switching Circuits .........
11.3.5
Improved Zero Voltage Triac Switches .......
11.3.6
Transistorized Zero Voltage Trigger ..........
1104
Use of GEL300 - A Monolithic Zero
Voltage Switch .............................
XIV
261
262
264
265
267
267
270
271
272
274
276
277
280
282
287
287
287
288
291
292
295
295
296
296
297
298
300
300
302
303
304
307
307
307
310
310
311
312
313
314
315
316
TABLE OF CONTENTS
11.4.1
11.4.2
11.4.3
11.5
11.6
Output and Power Connections. . . . . . . . . . . . ..
Mating the IC to the High Current SCR. . . . . ..
Connections for the Input Section. . . . .
Zero Voltage Switching at High Frequencies. . . . . .
Three Phase Zero Voltage Switching Power Control..
317
319
319
320
321
12. SOLID STATE TEMPERATURE AND AIR CONDITIONING CONTROL . ...... ' 325
12.1
12.2
12.2.1
12.2.2
12.2.3
12.2.4
12.2.5
12.3
12.4
12.4.1
12.4.2
12.5
12.5.1
12.5.2
12.5.3
12.5.4
12.5.5
12.5.6
12.5.7
12.5.8
12.6
.12.6.1
12.6.2
12.6.3
12.6.4
Introduction .................................
How to Select the Proper Control. . . . . . . . . . . . . .
Thermal System Model ........
Elements of Feedback Control
... .....
Phase Shifts
Proportional Control ...
Controller Specifications
Phase Control Vs Zero Voltage Switching.
Phase Control Circuits ...
Remote Sensor
. . . . . . . .. ....
Linear Phase Control
Zero Voltage Switching Circuits ................. ,
Zero Voltage Switching With the GEL300
.......... '
Integrated Circuit
Zero Voltage Switching With an Inductive Load
Proportional Control With Zero Voltage
Switching ............................
Low Power Zero Voltage Switching Using
the GEL300 ................... .
How to Use Low Resistance Sensors ..
Multiple Triac Triggering ....
Load Staging
........
Fail Safe Operation
Air Conditioning
.. ... .
. .................. .
Cooling .. .. ..
Ventilating
....................
Ventilating Blower Control for Heating and
Cooling ......................... .
Fan and Coil Blower Control ........... .
13. CHOPPERS, INVERTERS AND CYCLOCONVERTERS . . . . . . . . . . . . . . . .
13.1
13.1.1
13.1.2
13.1.3
13.1.4
13.1.5
13.2
13.2.1
13.2.1.1
13.2.1.2
13.2.2
13.2.2.1
Classification of Inverter Circuits ...........
Classes of Inverter Circuits ................
Properties of the Inverter Classes ............
Inverter Configurations ....... ......
Properties of the Different Inverter
Configurations ...... .. ........
Discussion of Classification System. . . .
Typical Inverter Circuits ... ..................
A Class A Inverter .. ..
. . . . . . . . . . . ..
Circuit Description ..................
Applications
..................
A Class B Inverter ........ . . . . . . . . . . . . . . ..
Circuit Description ................... ,
325
325
326
327
328
328
329
330
330
331
333
334
334
336
337
339
339
340
341
342
342
343
344
346
348
351
351
352
352
353
353
354
354
354
354
356
357
3.57
xv
SCR MANUAL
13.2.2.2
13.2.3
13.2.3.1
13.2.3.2
13.2.3.3
13.2.4
13.2.4.1
13.2.4.2
13.2.4.3
13.2.4.4
13.2.4.5
13.2.5
13.2.5.1
13.2.5.2
13.3
13.3.1
13.3.2
13.3.2.1
13.3.2.2
13.3.2.3
13.3.2.4
13.3.3
13.3.3.1
13.3.3.2
13.3.3.3
13.3.3.4
13.3.3.5
13.3.3.6
13.3.3.7
13.3.3.8
13.3.4
13.3.4.1
13.4
13.5
13.5.1
13.5.2
13.6
Circuit Performance ..................
Class C Inverters ..........................
Ott Filters for Class C Inverters ......... ;
Design Procedure ....................
A 400 Hz Inverter With Sine Wave Output
Designing a Battery Vehicle Motor-Controller
Using the Jones SCR Chopper (Class D) .....
Introduction .........................
Operation of the Jones Commutation Circuit
Design Trade-Offs ....................
Design Notes ........................
Worked Example .' . . . . . . . . . . . . . . . . . ..
Pulse Width Modulated (PWM) Inverter. . . . ..
The Auxiliary Commutated Inverter
(Class D) ..........................
Design Notes ........................
Inverter Accessories ., . . . . . . . . . . . . . . . . .'. . . . . . ..
The Ability to Operate Into InduCtive Loads. ..
Overcurrent Protection ..... :..............
Fuses and Circuit Breakers in the
DC Supply . . . . . . . . . . . . . . . . . . . . . . ..
Current Limiting by Pulse-Width Control..
Current Limiting by LC Resonance. . . . ..
Current Limiting in Class A Circuits by
Means of Series Capacitors. . . . . . . . . ..
Sine Wave Output ...................... "
Resonating the Load. . . . . . . . . . . . . . . . . ..
Harmonic Attenuation by Means of an
LC Filter .........................
An LC Filter Plus Optimum Pulse Width
Selection . . . . . . . . . . . . . . . . . . . . . . . . ..
Synthesis by Means of Output Voltage
Switching .........................
Synthesis by Controlling the Phase
Relationship of Multiple' Inverters. . . . ..
Multiple Pulse' Width Control. . . . . . .
Selected Harmonic Reduction. . . . . . .
The Cycloinverter .............
Regulated Output ....................
Supply-Voltage Regulation .........
Pulse Modulator Switches ......................
Cycloconverters .............................
Basic Circuit ...........................
Polyphase Application ... . . . . . . . . . . . . . . . . ..
Selected Bibliography .........................
359
361
362
364
367
369
369
371
375
377
380
383
383
385
387
388
389
389
389
390
390
391
392
392
392
393
394
395
395
396
396
396
397
397
398
399
399
14. LIGHT ACTIVATED THYRISTOR APPLICATIONS .. . . . . . . . . . . . . . . . . .. 409
14.1
14.1.1
14.1.2
14.1.3
14.1.4
XVI
Light Activated Semiconductors ....... ; .........
Photo Diode (Light Sensitive Diode) ..........
Photo Transistors .........................
Photo Darlington Amplifier. . . . . . . . . . . . . . . ..
Light Activated SCR (LASCR) ............ "
409
409
411
413
414
TABLE OF CONTENTS
14.1.5
14.2
14.2.1
14.2.2
14.3
14.3.1
14.4
14.4.1
14.4.2
14.4.3
14.4.4
14.4.5
14.4.6
14.5
14.5.1
14.5.2
14.5.3
14.5.4
14.5.5
14.5.6
14.5.7
14.5.8
14.5.9
14.5.10
14.5.11
14.5.12
14.6
14.6.1
14.6.2
14.6.3
Light Activated Silicon Controlled Switch
(LASCS) ......
. . . . . . . . . . . . . ..
Light Emitting Devices. . . .
. .............
Tungsten Lamps
...................
Light Emitting Diodes (LED) or Solid State
Lamps (SSL) ........................
Photon Coupler
... ......
..........
Specifications of Light Intensity. . . . . .
Characteristics of Sources and Sensors. . . . . . . . . . ..
Definition of Light Intensity. . . . . . . . .
Design Procedures
... ... .....
Effective Irradiance to Trigger. . . . . .
Approximate Irradiance Calculation.. . . .
Refined Irradiance Calculations
Comparison of Sources. . . . . . . . . . . . . . .
Applications ..........................
Light Activated DC and AC Relays ........ "
Trigger Higher Power SCR's by Light. . . . .
Light Activated Triac Applications. . . . . . . . . ..
Light Activated Integrated Zero Voltage Switch.
Light Activated Integrated Circuit Phase Control
Series Connection of SCR's Triggered by Light
(Light Activated High Voltage Switch). . . . ..
Light Activated Logic Circuits. . . . . . . . . . . . ..
Light Activated Astable CIrcuits. . . . . . . . . . . ..
Light Interruption Detector. . . . . . . . . . . . . . . ..
Higher Sensitivity Light Detectors. . . . . . . . . ..
"Slave" Electronic Flash ...................
Light Activated Motor Control. . . . . . . . . . . . ..
Circuits for Light Emitting Devices. . . . . . . . . . . . ..
Low Loss Brightness Control. . . . . . . . . . . . . . ..
Current Limiting Circuits ..................
Impulse Circuits for Light Emitters. . . . . . . .
15. PROTECTING THE THYRISTOR AGAINST OVERLOADS & FAULTS. . . . . . . ..
15.1
Why Protection? ..............................
15.2
Overcurrent Protective Elements. . . . . . . . . .
15.3
Coordination of Protective Elements. . . . . . . . . . . . ..
Protecting Circuits Operating on Stiff Power Systems
15.4
15.4.1
The Current Limiting Fuse. . . . . . . . . . . . . . . ..
Fuse-SCR Coordination in AC Circuits. . . . . . ..
15.4.2
15.4.2.1
Fuse Ratings .........................
15.4.2.2
SCR Rating for Fuse Application. . . . . . ..
15.4.2.3
Selecting a Fuse for SCR Protection. . . . ..
15.4.3
Fuse-SCR Coordination in DC Circuits. . . . . ..
15.5
Interrupted Service Type Fault Protection Without
Current Limiting Impedance .................
15.6
Non-Interrupted Service Upon Failure of Semiconductor .................................
15.7
Overcurrent Protection Using Gate Blocking .......
15.8
Overcurrent Protection Circuit ................. '.
418
419
419
423
425
426
426
427
428
429
430
430
431
432
432
433
434
435
436
437
438
439
440
440
441
442
442
443
444
445
447
447
448
449
450
451
451
453
458
458
459
461
464
466
466
XVII
SCR MANUAL
16. VOLTAGE
16.1
16.2
16.2.1
16.2.2
16.2.3
16.2.4
16.3
16.3.1
16.3.1.1
16.3.2
16.3.2.1
16..3.2.2
16.4
TRANSIENTS IN THYRISTOR CIRCUITS ..................
Where to Expect VoItageTransients ..............
How to Find Voltage Transients ................ ,
Meters. . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Oscilloscopes ........ " . . . . . . . . . . . . . . . . . ..
Peak Recording Instruments ............... ,
Spark Gaps ..............................
Suppre~sion Techniques .......................
Suppression Components . . . . . . . . . . . . . . . . . ..
Polycrystalline Suppressors .............
Suppression Network . . . . . . . . . . . . . . . . . . . . ..
Snubber Calculation for DC Circuit ..... '
Snubber Calculation for AC Circuit. . . . ..
Miscellaneous Methods ........................
17. RADIO FREQUENCY INTERFERENCE AND INTERACTION
OF THYRISTORS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Introduction. . . . . .......................... ,
17.1
17.2
The Nature of Radio Frequency Interference (RFI)..
17.2.1
Filter Design ...................... , .. ".
17.2.2
Components for R-F Filters, .... , .. , , .. , , ...
17.2.3
Fast Recovery Rectifiers ,., .. ,.".,'., ... ,.
17.2.4
Reduction of Radiated RFI .. ,." .. ,., ... ,'
17.2.5
Zero Voltage Switching .. , .. ,.,., ... ,' ... ,.
17.3
Interaction .,.,.. ...,"',...................
17.3.1
Interaction Acting on Anode Circuit ... , ... ' ..
17.3.2
Interaction Acting on the Trigger Circuit, , .. ,.
17.4
Decoupling the UJT Trigger Circuit Against
Supply Transients .,. "".".,., ... ,.......
17.5
Decoupling UJT Circuits Against SCR Gate
Transients ."...,..',..,.,..,.',....,....'
17.6
Good Design Practices to Minimize Sources of
SCR Interaction ., , .. , , , , . , , ,. , .. , ....... , . .
17.7
E.M.I. Standards and Restrictions .. , ............ ,
469
469
473
473
473
474
476
477
477
477
481
482
484
486
489
489
489
490
493
495
496
497
497
497
498
498
499
.499
500
18. MOUNTING AND COOLING THE POWER SEMICONDUCTOR .. , .' ...... ' ., 503
18.1
Lead Mounted SCR's .. ', ....... , .. , .... , .. ," 503
18.2
Mounting SCR's to Heat Exchangers. , . , ' .. , . . . .. 504
18.2.1
Case to Heat Exchanger Interface Considerations 504
18.2.1.1
Exchanger Surface Preparation, .... ' . , .. 505
18.2.1.2
Interface Thermal Grease, , , . , ... , ..... , 505
18.2.1.3
Electrical Isolation Case to Heat Exchanger 507
18.2.2
Mounting the Power Tab .. , ..... ,......... 511
18.2.3
Mounting the Power Pac Package (TO-220) ... , 514
18.2.4
Mounting the Press-Fit Package, ' , , ........ ,. 517
18.2.5
Mounting the Stud Type SCR. , . , ... ' . , . . . .. 518
18.2.6
Mounting the Flat Base Semiconductor, . . . . . .. 519
18.2.6.1
Heat Exchanger Thickness, , , .......... ' 520
18.2.6.2
Mounting Procedure .... , .. , ... , ...... ' 521
18.2.7
Mounting the Press Pak SCR, ... , ........ , .. 522
18.2.7.1
Mounting Clamp Requirements, .. , ..... , 523
18.2.7.2
Multiple Unit Mounting, . , . , ...... , ... , 527
XVIII
TABLE OF CONTENTS
18.2.7.2.1
18.2.7.2.2
18.2.8
18.2.8.1
18.2.8.2
18.3
18.3.1
18.3.1.1
18.3.1.1.1
18.3.1.1.2
18~3.1.1.3
18.3.1.1.4
18.3.1.1.5
18.3.1.1.6
18.3.1.2
18.3.1.3
18.3.2
18.3.'2.1
18.3.2.2
18.3.2.3
18.3.2.2.1
18.3~2.2.2
18.4
18.4.1
18.4.2
Parallel
Series ......................... .
Unit Pak Mounting ....................... .
Preparation of Heat Exchanger ......... .
............. .
Mounting Procedure
Selecting a Heat Exchanger. . . . . . . . . . . . . . . . . . . . .
Low to Medium Current SCll's ............. .
Designing the Flat Fin Heat Exchanger .. .
General ........................ .
Radiation ...................... .
Free or Natural Convection. . . . . . . . .
Forced Convection .............. .
Fin effectiveness ................ .
Typical Example of Complete Fin
Design ..................... .
Example of Calculating the Transient
Thermal Impedance Curve for a Specific
Heat Exchanger Design. . . . . . . . . . . . ..
Selection of Commercial Heat Exchangers.
Medium to High Current SCR's. . . . . . . . . . .
Press Pak Vs Stud. . . . . . . . . . . . . . . .
Free Vs Forced Air Convection. . . . . . . . ..
Liquid Cooling . .
. . . . . . . . . . . . . . . . ..
Heat Exchanger Selection. . . . . . . . ..
Liquid Selection and Requirements. ..
Measurement of Case Temperature. . . . . . . . . . . . . ..
Materials Used .' . . . . . . . . . . . . . . . . . . . . . . . ..
Preparation
. . . . . . . . . . . . . . . . . . . . . . ..
527
528
529
529
530
530
530
530
530
532
533
535
536
539
540
541
542
542
543
546
546
548
549
550
551
19. SCR RELIABILITY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 553
19.1
19.2
19.3
19.3.1
19.4
19.5
19.6
19.6.1
19.6.2
19.6.3
19.6.4
19.6.5
19.6.6
19.7
19.8
..............................
Introduction
What is Reliability .......................... "
Measurement of Reliability. . . . . . . . . . . . . . . . . . . ..
Failure Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
SCR Failure Rates ............................
Designing SCR's for Reliability ..................
Failure Mechanisms ...........................
Structural Flaws .........................
Encapsulation Flaws ......................
Internal Contaminants .....................
Material Electrical Flaws. . . . . . . . . . . . . . . . . ..
Metal Diffusion ..........................
Nuclear Radiation ........................
Effects of Derating ............................
Reliability Screens ............ . . . . . . . . . . . . . . ..
553
553
554
554
555
555
557
557
558
558
559
559
559
560
563
20. TEST CIRCUITS FOR THYRISTORS . . . . . . . . . . . . . . . . . . . . . . . . . . .. 565
20.1
20.2
20.3
Introduction ................................. 565
Instrumentation .............................. 565
Specified Peak Off-State and Specified Reverse
Voltage ................................... 567
XIX
SCR MANUAl
20.3.1
20.4
20.5
20.5.1
20.5.2
20.5.3
20.5.4
20.6
20.7
20.8
20.8.1
20.8.2
20.9
20.10
20.10.1
20.11
20.11.1
20.11.2
20.12
20.13
20.14
20.14.1
.20.15
20.15.1
20.15.2
20.15.3
20.16
20.17
Specified Peak Off-State and Specified
Reverse Voltage for Thyristors. . . . . . . . . . ..
Peak Reading Voltmeter ...................... "
DC-Gate Trigger Current and Voltage Test ...... "
Anode Supply for Gate Test. . . . . . . . . . . . . . . ..
DC Gate Supply for Gate Test. . . . . . . . . . . . . ..
Pulse Gate Supply for Gate Test .............
Gate Trigger Test Set for Low Current SCR's
(Less Than 2 Amperes Current Rating). . . ..
DC Holding Current Test. . . . . . . . . . . . . . . . . . . . . ..
Latching Current Test. . . . . . . . . . . . . . . . . . . . . . . ..
Peak On-State Voltage Test Circuit. . . . . . . .. . . . . ..
On-State Voltage (Low Level) (25°C) ....... "
On-State Voltage (High Level) ............ "
Critical Rate of Rise of On-State Current Test (dildt)
Turn-On Voltage Test .........................
Gate Controlled Turn-On Time. . . . . . . . . . . . ..
Dv/dt Test - Critical Rate of Rise of Off-State
Voltage Test. .. . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Exponential dv/dt Test .....................
Linear dv/dt Test .........................
Critical Rate of Rise of Commutating Off-State
Voltage for Bidirectional Thyristors (Triacs) Test ..
Turn-Off Time Test ........ . . . . . . . . . . . . . . . . ..
Thermal Resistance Test. . . . . . . . . . . . . . . . . . . . . . ..
Thermal Resistance of Press Pak Rectifier
Diodes & Thyristors. . . . . . . .. . . . . . . . . . . ..
Testing Thyristors on Curve Tracers ............ "
Off-State & Reverse Voltage ............... "
Gate Voltage, Gate Current Measurement. . . ..
Forward Current & On-State Voltage
Measurement ..........................
Elevated Temperature Testing. . . . . . . . . . . . . . . . . ..
Commercial Thyristor Test Equipment. . . . . . . . . . ..
21. SELECTING THE PROPER THYRISTOR AND CHECKING THE
COMPLETED CIRCUIT DESIGN ............................. .
21.1
Selecting the Proper Thyristor. . . . . . . . . . . . . . . . . ..
21.1.1
Semiconductor Design Trade-Offs ........... .
21.1.2
Selection Check List. . . . . . . . . . . . . . . . . ..... .
21.2
Checking Circuit Design ...................... .
21.2.1
Thyristor Ratings & Characteristics. . . . . . . . .. .
21.2.2
Voltage Measurement ..................... .
21.2.3
Current Measurement ., .................. .
21.2.4
The Power Circuit ...................... .
21.2.5
Modifications to Soften dvI dt. . . . . . . . . . . . . ..
21.2.6
Modifications to Soften Initial dil dt. . . . . . . . ..
21.2.7
Gate Circuit ............ : ............... .
21.2.8
Temperature Measurement ................ .
21.2.9
Magnetic Saturation ...................... .
21.2.10
Supply Impedance ....................... .
21.3
SCR Selection Examples ....................... .
xx
567
569
570
571
572
573
574
576
578
579
579
580
582
583
584
585
586
586
588
589
591
595
595
596
596
597
598
598
599
599
599
600
601
602
602
602
604
604
604
604
606
606
607
607
TABLE OF CONTENTS
21.3.1
21.3.2
21.3.3
21.3.4
21.4
Current Conversion Factors. . . . . . . . . . . . . . . ..
Definition of Terms. . . . . . . . . . . . . . . . . . . . . . ..
SCR Bridge .............................
Inverter SCR Selection. . . . . . . . . . . . . . . . . . . ..
Check List ..................................
607
607
610
613
614
22. GENERAL ELECTRIC THYRISTOR AND DIODE CONDENSED
SPECIFICATIONS ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
22.1
Phase Control SCR's. . . . . . . . . . . . . . . . . . . . . . . . . ..
22.2
Inverter SCR's ...............................
22.3
Triacs ......................................
22.4
Triac Trigger Devices. . . . . . . . . . . . . . . . . . . . . . . . ..
22.5
Optoelectronic Devices ...................... ..
22.6
Silicon Rectifiers .............................
22.7
Circuit Assemblies ... . . . . . . . . . . . . . . . . . . . . . . . ..
22;8
Rectifier & SCR Modules. . . . . . . . . . . . . . . . . . . . . ..
22.9
Selenium Components ............ . . . . . . . . . . . ..
22.10
GE-MOV Metal Oxide Varistors. . . . . . . . . . . . . . . ..
615
616
623
629
633
638
642
648
651
654
656
23. APPLICATION AND SPECIFICATION LITERATURE; SALES OFFICES. . . . . . ..
23.1
Semiconductor Device Catalogs. . . . . . . . . . . . . . . . ..
23.2
Application Notes ............................
23.2.1
General Applications for Power Semiconductors.
23.2.2
Silicon Controlled Rectifier and Other Thyristor
Circuits ............................. ..
23.2.3
Unijunction Applications .. . . . . . . . . . . . . . . . ..
23.2.4
Test Circuits .. . . . . . . . . . . . . . . . . . . . . . . . . . ..
23.3
Specification Sheets ... . . . . . . . . . . . . . . . . . . . . . . ..
23.4
Related General Electric Departments. . . . . . . . . . . ..
23.5
General Electric Sales Offices. . . . . . . . . . . . . . . . . . ..
23.6
International GE Sales Offices. . . . . . . . . . . . . . . . . ..
23.7
General Electric Semiconductor Distributors. . . . . ..
Application Index .......................................
Index .................................................
657
657
657
658
659
659
659
660
660
660
666
666
671
674
XXI
SCR MANUAL
XXII
CONSTRUCTION AND BASIC THEORY OF OPERATION
1
CONSTRUCTION AND BASIC THEORY OF OPERATION
1.1 WHAT IS A THYRISTOR?
The name thyristor l defines any semiconductor switch whose
bistable action depends on p-n-p-n regenerative feedback. Thyristors
can be two, three, or four terminal devices, and both unidirectional
and bi-directional devices are available.
1.2 CLASSIFICATIONS OF THYRISTORS
The silicon controlled rectifier (SCR) is by far the best known of
all thyristor devices. Because it is a unidirectional device (current flows
from anode to cathode only) and has three terminals (anode, cathode
and control gate), the SCR is classified as a reverse blocking triode
thyristor. Other members of the reverse blocking triode thyristor family
include the silicon unilateral switch (SUS), the light activated silicon
controlled rectifier (LASCR), the complementary SCR (CSCR), the gate
turn-off switch (GTO) , and the programmable unijunction transistor
(PUT). The silicon controlled switch (SCS) is a reverse blocking tetrode
thyristor (it has two control gates), while the Shockley diode is a
reverse blocking diode thyristor. Bidirectional thyristors are classified
as p-n-p-n devices that can conduct current in either direction; commercially available bidirectional triode thyristors include the triac (for
triode AC switch), and the silicon bilateral switch (SBS).
1.3 TWO TRANSISTOR ANALOGY OF p-n-p-n OPERATION 2
A simple p-n-p-n structure, like the conventional SCR, can best
be visualized as consisting of two transistors, a p-n-p and an n-p-n interconnected to form a regenerative feedback pair as shown in Figure 1.1.
ANOOE
p
GATE
p
N
•
-
IG(p)
CATHODE
FIGURE 1.1
-
IG(nl
p
TWO TRANSISTOR ANALOGUE OF P·N·P·N STRUCTURES
SCR MANUAL
From this figure, it is evident that the collector of the n-p-n transistor
(along with possible n-gate drive) provides base drive for the p-n-p
transistor.
IBl
= IC2 + IG(n)
(1.1)
Similarly, the collector of the p-n-p transistor along with any p-gate
current (IG(p») supplies the base drive for the n-p-n transistor.
(1.2)
Thus, a regenerative situation exists when the positive feedback gain
exceeds one.
,Ic'
_4_
II
IE + ICBO
'T'CCB
----""
FIGURE 1.2 COMMON BASE CURRENT RELATIONSHIPS
The p-n-p-n structure may be analyzed in terms of its common
base current gains (ap and an) and the avalanche multiplication coefficients of holes and electrons, Mp and Mn respectively. From 1.2, the
base current for the n-p-n transistor is seen to be:
.
IB2
= IK (1:- an Mn) -
(1.3)
I CBO (2)
where I CBO (2) is the collector-to-base leakage current of transistor Q2.
However, the collector current of the p-n-p is:
(104)
But since
IB2 = ICl + IG(p)
IA + IG(p) = IK + IG(n)
(1.5)
(1.6)
Equations 1.3 to 1.6 can be solved for
IA = an Mn IG(p)
Call "ap Mp
+ IG(n)
(1 - ap Mp) + ICBO(l)
1-ap Mp-an Mn
+ IcBo (2)
(1.7)
+ an Mn" the loop gain G.
With proper bias applied (positive anode to cathode voltage) and
in the absence of any gate signal, Mn and Mp are approximately unity
and the transistor alphas are both low. The denominator of Equation
1.7 approaches one, and IA is little higher than the sum of the individual transistor leakage currents. Under these conditions the p-n-p-n
structure is said to be in its forward blocking or high impedance "off"
state. The switch to the low impedance "on" state is initiated simply
by raising the loop gain to unity. Inspection of Equation 1.7 shows that
as this term ~ I; IA ~ co. Physically, as the loop gain approaches
unity and the circuit starts to regenerate, each transistor drives its mate
2
CONSTRUCTION AND BASIC THEORY OF OPERATION
into saturation. Once in saturation, all junctions assume a forward bias,
and total potential drop across the device approximates that of a single
p-n junction. Anode current is limited only by the external circuit .
.
1-------------------
FIGURE 1.3
EMlnER CURRENT DEPENDENCE OF
a;
IN A SILICON TRANSISTOR
The loop gain G can approach one (1) either by an increase in
Mp or Mn with increasing voltage or with an increase in the alphas
with either voltage or current. In most silicon transistors, C% is quite
low at low emitter currents, but increases fairly rapidly as emitter current is increased. This effect (Figure 1.3) is due to the presence in the
silicon of special impurity centers. Any mechanism which causes a
temporary increase in transistor emitter current is therefore potentially
capable of turning on a p-n-p-n device. The most important of these
mechanisms are:
1. Voltage. As the collector-to-emitter voltage of a transistor is
increased, eventually a point is reached where the energy of the (leakage) current carriers arriving at the collector junction is sufficient to
dislodge additional carriers. These carriers in turn dislodge more carriers, and the whole junction goes into a form of avalanche breakdown
characterized by a sharp increase in collector current. In a p-n-p-n
device, when the avalanche current makes G ~ 1, switching takes
place. This is the turn-on mechanism normally employed to switch four
layer diodes into conduction.
2. Rate of change of voltage. Any p-n junction has capacitance the larger the junction area, the larger the capacitance. Figure 1.2
shows the collector-to-base capacitance dotted in. If a step function of
voltage is impressed suddenly across the collector-to-emitter terminals
of the transistor, a charging current i will How from the emitter-tocollector to charge the device capacitance
i
= C dv/dt
(1.8)
In Figure 1.1, the charging current Howing in the p-n-p represents
base current for the n-p-n so that G can rapidly approach 1, switching
on the device.
3. Temperature. At high temperatures, leakage current in a reverse biased silicon p-n junction doubles approximately with every 8°C
increase in junction temperature. When the temperature generated
3
5CR MANUAL
leakage current in a p-n-p-n structure has risen sufficiently for G ~ 1,
switching occurs.
4. Transistor Action. Collector current is increased in conventional transistor manner by temporarily injecting additional ("gate")
current carriers into a transistor base region. This is the machanism
normally employed to tum-on SCR's and other p-n-p-n devices which
have an external connection ("gate" lead) to one or more of the transistor bases.
By convention, SCR's that are turned on by injecting current into
the lower p-base (via an external connection to this base) are called
"conventional" SCR's, while those that withdraw current from the upper
n-base are called "complementary" SCR's. The four-terminal silicon
controlled switch (SCS) has connections made to both bases and either
or both bases may be used to initiate switching. More external gate
current is required to trigger a p-n-p-n device via the n-gate rather
than the p-gate for two reasons:
a) The uppern-base region is of high resistivity silicon (the upper
p-n junction supports the major part of any applied reverse
voltage) so that ilp is very small.
b) From Figure 1.1, n-base drive necessarily removes current from
the loop and therefore reduces G.
5. Radiant energy ("light"). Incident radiant energy within the
spectral bandwidth of silicon impinging on and penetrating into the
silicon lattice releases considerable numbers of hole electron pairs. When
the resultant device leakage current climbs above the critical level for
G ~ 1, triggering will ensue. This triggering mechanism makes possible the light activated SCR. In these devices a translucent "window"
is provided in the device encapsulation in order that "light" may reach
the silicon pellet. The LASCR, because it is provided with a gate
lead, may be triggered either by light or by electrical gate current.
Chapter 14 is devoted to light activated thyristors.
1.4 REVERSE BLOCKING THYRISTOR (SCR) TURN-OFF MECHANISM
When a reverse blocking thyristor is in the conducting state, each
of the three junctions of Figure 1.4 are in a condition of forward bias
and the two base regions (B p , Bx) are heavily saturated with holes and
electrons (stored charge).
Ep
BN
Bp
EN
P
N
P
N
CATHODE
ANODE
IT
r
JI
J2
J3
FIGURE 1.4 THYRISTOR BIASED IN CONDUCTING STATE (GATE OPEN CIRCUITED)
4
CONSTRUCTION AND BASIC THEORY OF OPERATION
To turn-off the thyristor in a minimum time, it is necessary to
apply a reverse voltage. When this reverse voltage is applied the holes
and electrons in the vicinity of the two end junctions (Jl, Ja) will diffuse
to these junctions and result in a reverse current in the external circuit.
The voltage across the thyristor will remain at about +0.7 volts as long
as an appreciable reverse current flows. After the holes and electrons in
the vicinity of 11 and 13 have been removed, the reverse current will
cease and the junctions I J and Ja will assume a blocking state. The
reverse voltage across the thyristor will then increase to a value determined by the external circuit. Recovery of the device is not complete,
however, since a high concentration of holes and electrons still exists
in the vicinity of the center junction (h). This concentration decreases
by the process of recombination in a manner which is largely independent of the external bias conditions. After the hole and electron
concentration at h has decreased to a low value, J2 will regain its blocking state and a forward voltage (less than V(BO») may be applied to the
thyristor without causing it to tum-on. The time that elapses after the
cessation of forward current flow and before forward voltage may safely
be reapplied is called the thyristor "turn-off time," tq and can range
from several microseconds to as high as several hundred microseconds
depending upon the design and construction of the particular thyristor.
1.5 IMPROVEMENTS FOR DYNAMIC SCR OPERATION
Ever since its introduction, circuit design engineers have been
subjecting the SCR to increasing levels of operating stress and demanding that these devices perform satisfactorily there. Different stress
demands that the SCR must be able to meet are:
1) Higher blocking voltages
2) More current carrying capability
3) Higher di/dt's
4) Higher dv/dt's
5) Shorter turn-off times
6) Lower gate drive
7) Higher operating frequencies
There are many different SCR's available today, which can meet one
or more of these requirements. But as always, an improvement in one
characteristic is usually only gained at the expense of another. It is
constructive to consider five device attributes which in combination
determine the thyristor's current rating and switching capability.
1. Voltage. There are four methods to increase the voltage rating
- surface contouring, increasing the Bll base width and/or its resistivity,
and/or its lifetime.
a) Surface contouring or beveling allows higher voltage operation
by reducing the electric fields at the surface of the silicon pellet. It is the most advantageous since it achieves this with no
other sacrifice in SCR operation except for a slight current
derating due to decreased emitter area. Beveling techniques
generally are used only on premium-type industrial devices,
5
SCR MANUAL
\'vherc the added costs due to a bigger silicon pellet plus the
requirement that the pellet be round for maximum beveling
effectiveness can be justified. Consumer type devices whose
pellets generally are "scribed" into square or rectangular shapes
for lowest cost are generally not beveled. Beveling effectiveness
here is destroyed by field concentration effects at the pellet
corners.
b) The most obvious way to increase the voltage· rating is to
increase the base width Bn (see Section 1.4) or its resistivity.
However, these actions will also increase the "on-state" voltage
drop so that the allowable current decreases. High frequency
performance is degraded because the di/dt rating goes down
and turn-off time increases due to increased stored charge in
this base.
c) A more subtle way is td increase the minority carrier lifetime
in Btl. This decreases the thermally generated leakage current
in the depletion region (IcBo).AdditionaI benefits are lower
"on-state" voltage and better di/ 9.t rating. Of courSe, turn-off
time will also increase and dv / dt withstand capability decreases
(see also Section 5 below, Turn-Off Time).
2. Current. The allowable current through a device depends primarily upon the "on-state" voltage. Any variable that decreases this
voltage drop (and hence, internal power dissipation), will raise the cur·rent limit. Favorable effects are larger pellets, smaller Bn base width,
higher lifetimes, and lower silicon resistivity.
3. di/dt. The problem of di/dt failures is well recognized today.3
Essentially the SCR is trying to conduct too much current through too
small a pellet area during the initial turn-on and the resultant, localized
junction temperature rise burns it out. The obvious answer is to have
the SCR turn on more area initially and this is precisely the objective
of the newer gate structures discussed below.
a) Conventional Side Gate or Point Gate (Eigure_ 1.5(a)). This is
inherently the simplest gate structure and provides adequate
performance for low di/dt applications. The area initially
turned on is quite small and depends upon the amplitude of
the gating signal. All of the early SCR's had a point contact
gate and it is still the most common gate amongst the consumer
and light industrial SCR's.
b) Conventional Center Gate (Figure 1.5(b)). Same as the conventional side gate except the di/ dt ratings are generally higher
due to the fact that a small circle rather than a point is turned
on. Di/dt capability is still quite a strong function of gate drive.
c) Field Initiated (F.I.) Gate (Figure 1.5(c)). This gate structure
is designed to turn on a definite length of SCR emitter periphery even with soft gate drive. The F.1. gate is a double switching type gate. That is, from an equivalent circuit view, it
behaves as a main SCR triggered by a small pilot SCR. A
portion of the anode current of the pilot SCR becomes the
gate drive to the main~ By this technique, only a small amount
of gate drive is required to li)itiate conduction in the pilot portion of the SCR. With this type of structure, delay time is a
6
CONSTRUCTION AND BASIC THEORY OF OPERATION
"l-
G1+ _
N
I...
II
p
N
p
.1+
FIGURE 1.5(a)
CONVENTIONAL SIDE GATE SCR
FIGURE 1.5(b)
CONVENTIONAL GATE SCR
FIGURE 1.5(d)
N+ GATE SCR
FIGURE 1.5(c)
F·1 GATE SCR
FIGURE 1.5(e)
JUNCTION DIAGRAM, PLAN VIEW AND EQUIVALENT CIRCUIT FOR
AMPLIFYING GATE SCR
7
SCR MANUAL
FIGURE 1.5(f) THE INTERDIGITATED AMPLIFYING GATE STRUCTURE IS SHOWN ON THE
LEn COMPARED TO THE CIRCULAR AMPLIFYING GATE
strong function of gate drive. Even though high di/dt is
achieved with soft drive, the circuit designer may be forced to
provide a stiff gate drive to achieve relatively short delay time.
d) N+ Gate (Figure 1.5(d)). This gate structure is designed to
turn-on a definite length of line, but not with soft gate drive;
here a stiff drive is required. However, with this structure a
consistently short delay time is achieved. This enhanced parameter along with others makes SCR's fabricated with the N+
gate ideal for applications requiring series and/or parallel
operation of SCR's.
e) Amplifying Gate (Figure 1.5(e)). The principals involved in
the amplifying gate are quite similar to those employed in the
F.r. structure. That is, double switching is employed but optimized to the point where the main SCR portion switches immediately after conduction is initiated in the pilot portion of the
device. The anode current of the pilot portion of the device
causes rapid turn-on of a significant portion of the main currentcarrying portion of the device. The three most important criteria
to building an optimized amplifying gate SCR are:
1. Complete turn-on of the emitter periphery of the main
current-carrying area to insure low switching loss density.
2. Instantaneous turn-on of the main current-carrying area
after turn-on of the pilot area.
3. Complete turn-on of the pilot portion of the device.
The design of the amplifying gate structure optimizes these
three criteria. To initiate turn-on, only a relatively small amount
of conventional gate current is required due to the pilot SCR
action - thus the term "amplifying." Schematically, the amplifying gate SCR looks like a main power handling SCR triggered
by a small "pilot" SCR. The area of immediate turn-on is sizeable as shown on the plan-view of the pellet structure as shown
in Figure 1.5(e).
8
CONSTRUCTION AND BASIC THEORY OF OPERATION
f) Distributed Amplifying Gate (Figure 1.5(f)). Although it is
possible to tum on a ring around the amplifying gate as previously described, a finite spreading velocity requires a finite
period of time before conduction moves out from the periphery
of the amplifying gate to cause the entire annular portion of
the thyristor to be "on." Depending on the designed voltage
and speed characteristics of a thyristor, the spreading velocity
can range anywhere from 3000 to 8000 cm/second. 5 This represents spreading times of 50 ,...seconds for a lIO ampere device
to better than 300 p.seconds for a 550 ampere device. One can
appreciate that for narrow pulses (less than the above times)
something less thim the entire pellet is in conduction and wide
pulse current capability cannot be realized.
The solution employed to extend the amplifying gate
principle to thyristors designed for narrow pulse application is
interdigitation. Figure 1.5(f) shows the extension of the amplifying gate technology to include interdigitation. The fingers
extending out into the cathode area increase the gate periphery
from something a little better than 2 em to about 33 centimeters. The distance now which spreading must traverse is
much reduced so that total tum-on occurs in something on the
order of 10 to 30 p.seconds. Of course the region for conduction
is much reduc~d with interdigitation but "who cares?"; you
wouldn't use any more for narrow pulses and you wouldn't use
an interdigitated device for wide pulses. The previously mentioned non-interdigitated devices are much superior for wide
pulse applications.
4. dv/dt. It was mentioned in Section 1.3 that a rapidly rising
voltage waveform could switch on an SCR. Since this can lead to spurious operation, SCR's to be used in circuits With high dv/dt's should be
of the "shorted emitter" construction with its intrinsic high dv/dt withstand capability.9
Figure 1.6 shows a "shorted emitter" thyristor structure. Externally applied gate current IG Hows from gate to cathode lateraUy
through the gate p-region. The voltage drop developed across the lateral
base resistance of p. forward biases the right hand edge of the cathode
junction. If gate current is sufficiently large, electrons are injected from
this point, and the device turns on in normal p-n-p-n fashion when
regeneration begins.
K
,..-!!---,...--...,
" SHORTED"
CATHODE
REGION I
I
•
ELECTRON
FLOW
!
CURRENT
FLOW
GATE
.I-++--<>G
N
p
ANODE
FIGURE 1.6 SHORTED EMlnER STRUCTURE
9
SCR MANUAL
The effect of the partial gate to cathode short is the same asplacing a resistor in paraliei with the gate cathode junction of a conventional non-shorted emitter device. 2 This resistor (RG in Figure 1.6)
diverts some of the thyristor's thermally generated leakage current
and/or dv / dt induced capacitive charging current around the gate-cathode junction, by providing an alternative lower impedance path to the
cathode. Regeneration is reduced and a shorted emitter thyristor has,
as a result, superior high temperature characteristics and dv/ dt capability. Emitter shorts reduce the emitter area that can conduct principle
current and also interfere with the tum-on of the device, thereby
reducing di/dt. 5
5. Tum-off Time (tq ). Section 1.4 pointed out that the stored
minority charge in the Bn base must decay to zero by recombination
before forward voltage may be applied to the SCR without its turning
on. This recombination effect can be represented by the simple formula
dPn
-
dtrecomb -
Pn
Tp
(1.9)
where Pn is the excess minority charge (holes in this case) and Tp is the
. lifetime of holes in the Bn base. Representative values of Tp range from
0.1 to 1000 ,...seconds and depend upon purity, structure and doping of
the silicon. High frequency operation requires short turn-off times so
that Tp must be made small. The usual way is by gold doping, i.e. the
introduction of gold impurities into Bm which act as additional recombination centers. However, as mentioned before, as Tp goes down, so
does the voltage and current ratings. Ial
~ +'~
NPN
Ial
Vc+
W +~
DARLINGTON AMPLIFIER
Ial
+t[.,
+t[+.
VC+
LIGHT SENSITIVE TRANSISTOR
(PHOTO TRANSISTOR)
.'~ '.
VC+
LIGHT SENSITIVE DARLINGTON
PHOTO AMPLIFIER
~'.
VC+'
UJT (UNIJUNCTION TRANSISTOR)
(N-TYPE BASE)
FIGURE 2.1
24
T
~-k
SEMICONDUCTOR GRAPHICAL SYMBOLS
IE
SYMBOLS AND TERMINOLOGY
NAME OF
SEMICONDUCTOR
DEVICE
GRAPHICAL SYMBOLS
USED IN
THIS MANUAL
Transistors (cont.)
CUJT (COMPLEMENTARY UNIJUNCTION
TRANSISTOR)
(P-TYPE BASE)
MAIN TERMINAL
V-I
CHARACTERISTIC
T ~
-V£-BI
~
BIDIRECTIONAL
TRIGGER DIAC (NPN TYPE)
THYRISTORS
PUT (PROGRAMMABLE UNIJUNCTION
TRANSISTOR)
~
~
;::LAPUT (LIGHT ACTIVATED PROGRAMMABLE
UNIJUNCTlON TRANSISTOR)
DIAC (BIDIRECTIONAL DIODE THYRISTOR)
K
-@--o
+
+
VA
~
VA+
+L__
1
~
SBS (SILICON BILATERAL SWITCH)
T
ASBS (ASSYMMETRICAL SILICON
BILATERAL SWITCH)
T
+
+-
SUS (SILICON UNILATERAL SWITCH)
FIGURE 2.1
SEMICONDUCTOR GRAPHICAL SYMBOLS
VA+
25
SCR MANUAL
NAME OF
SEMICONDUCTOR
DEVICE
GRAPHICAL SYMBOLS
USED IN
THIS MANUAL
Thyr.istor. (cont.)
SCR (SILICON CONTROLLED RECTIFIER)
REVERSE BLOCKING TRIODE
THYRISTOR
LAS (LIGHT ACTIVATED SWITCH)
LIGHT ACTIVATED REVERSE
BLOCKING DIODE THYRISTOR
~
oA
LASCR (LIGHT ACTIVATED SILICON
CONTROLLED RECTIFIER)
LIGHT ACTIVATED REVERSE
BLOCKING TRIODE THYRISTOR
~
Ko
~
~
TRIAC (BIDIRECTIONAL TRIODE
THYRISTOR)
SCS (SILICON CONTROLLED SWITCH)
REVERSE BLOCKING TETRODE
THYRISTOR
MAIN TERMINNV-I
CHARACTERISTIC
+
+
+
+
.
-
-,
VA+
VA+
VA+
V
4r ~
4r +
A
K
.
VA+
GI
ASCS (LIGHT ACTIVATED SILICON
CONTROLLED SWITCH)
LIGHT ACTIVATED REVERSE
BLOCKING TETRODE THYRISTOR
A~
K
-- .VA+
61
A-ANODE
B - BASE
C - COLLECTOR
E-EMITTER
6 - GATE
K - CATHODE
NOTE:
CIRCLES AROUND GRAPHICAL SYMBOLS ARE OPTIONAL EXCEPT
WHERE OMISSION WOULD RESULT IN CONFUSION. IN THESE CASES
CIRCLE DENOTES AN ENVELOPE THAT EITHER ENCLOSES A NONACCESSIBLE TERMINAL OR TIES A DESIGNATOR INTO SYMBOL.
FIGURE 2.1
26
SEMICONDUCTOR GRAPHICAL SYMBOLS
SYMBOLS AND TERMINOLOGY
2.2 SCR TERMINOLOGY
The following tabulation defines the terminology used in SCR
and triac specifications. As in the case of graphical symbols (Section
2.1) we try to conform to existing standards wherever possible.
2.2.1 Subscripts
The following letters are used as qualifying subscripts for thyristor
letter symbols.
A
(AV)
(BO)
(BR)
C
D
d
G
H
K
L
M
o
q
R
(RMS)
r
rr
S
T
e
W
Anode, Ambient
Average Value
Breakover
Breakdown
Case
Off-State, N on-Trigger
Delay
Gate
Holding
Cathode
Latching
Maximum Value
Open Circuit
Tum-off
Reverse or, as a second subscript, Repetitive
Total Root Mean Square Value
Rise
Reverse Recovery
Short Circuit, or as a Second Subscript, Non-Repetitive
(Infrequent)
On-State, Trigger
Thermal
Working
2.2.2 Characteristics and Ratings
A characteristic is an inherent and measurable property of a
device. Such a property may be electrical, mechanical, thermal, hydraulic, electro-magnetic or nuclear and can be expressed as a value for
stated or recognized conditions. A characteristic may also be a set of
related values, usually shown in graphical form.
A rating is a value which establishes either a limiting capability
or a limiting condition (either maxima or minima) for an electronic
device. It is determined for specified values of environment and operation, and may be stated in any suitable terms.
Principal Voltage-Current
Characteristic (Principal
Characteristic)
The function, usually represented graphically, relating the principal voltage to
the principal current with gate current,
where applicable, as a parameter.
27
SCR MANUAL
Anode-to-Cathode VoltageCurrent Characteristic
(Anode Characteristic)
A function, usually represented graph~
ically, relating the anode-to-cathode voltage to the principal current with gate
current, where applicable, as a parameter.
NOTE: This term does not apply to bidirectional thyristors.
On-State
The condition of the thyristor corresponding to the low-resistance, low-voltage
portion of the principal voltage-current
characteristic in the switching quadrant(s).
Off-State
The condition of the thyristor corresponding to the high-resistance, low-current
portion of the principal voltage-current
characteristic between the -origin and the
breakover point(s) in the switching quadrant(s).
Breakover Point
Any point on the principal voltage-current
characteristic for which the differential
resistance is zero and where the principal
voltage reaches a maximum value.
Negative Differential
Resistance Region
Any portion of the principal voltagecurrent characteristic in the switching
quadrant(s) within which the differential
resistance is negative.
Reverse Blocking State
(of a Reverse Blocking
Thyristor)
The condition of a reverse blocking
thyristor corresponding to the portion of
the anode-to-cathode voltage-current
characteristic for reverse currents of lower
magnitude than the reverse breakdown
current.
Off-Impedance
The differential impedance between the
terminals through which the principal
current Hows, when the thyristor is in the
off-state at a stated operating point.
On-Impedance
The differential impedance between the
terminals through which the principal
current Hows, when the thyristor is in the
on-state at a stated operating point.
The differential· impedance between the
Reverse Blocking Impedance (of a Reverse Blocking two terminals through which the principal current Hows, when the thyristor is in
Thyristor)
the reverse blocking state at a stated
operating point.
28
SYMBOLS AND TERMINOLOGY
Principal Voltage
Anode-to-Cathode Voltage
(Anode Voltage)
Forward Voltage (of a
Reverse Blocking Thyristor)
Off-State Voltage
Working Peak Off-State
Voltage
Repetitive Peak Off-State
Voltage
Non-Repetitive Peak
Off-State Voltage
Critical Rate of Rise of
Off-State Voltage
Reapplied Rate of Rise of
Voltage, Reapplied dv/dt
(of Reverse Blocking
Thyristor)
The voltage hetween the main terminals.
NOTE: 1. In the case of reverse blocking thyristors, the principal
voltage is called positive
when the anode potential is
higher than the cathode potential, and called negative
when the anode potential is
lower than the cathode potential.
2. For hi-directional thyristors,
the principal voltage is called
positive when the potential
of main terminal 2 is higher
than the potential of main
terminal 1.
The voltage hetween the anode terminal
and the cathode terminal.
NOTE: 1. It is called positive when the
anode potential is higher than
the cathode potential, and
called negative when the
anode potential is lower than
the cathode potential.
2. This term does not apply to
hi-directional thyristors.
A positive anode-to-cathode voltage.
The principal voltage when the thyristor
is in the off-state.
The maximum instantaneous value of the
off-state voltage which occurs across a
thyristor, excluding all repetitive and
non-repetitive transient voltages.
The maximum instantaneous value of the
off-state voltage which occurs across a
thyristor, including all repetitive transient voltages, hut excluding all nonrepetitive transient voltages.
The maximum instantaneous value of any
non-repetitive transient off-state voltage
which occurs across the thyristor.
The minimum value of the rate of rise of
principal voltage which may cause switching from the off-state to the on-state.
Rate of rise of forward voltage following
turn-off, or commutation. (This is a test
condition for tum-off time measurement.)
29
SCR MANUAL
Critical Rate of Rise of
Commutation Voltage (for
Bidirectional Thyristors)
The mInImum value of the rate of rise
of principal voltage which may cause
switching from the off-state to the onstate immediately following on-state current conduction in the opposite quadrant.
Breakover Voltage
The principal voltage at the breakover
point.
On-State Voltage
The principal voltage when the thyristor
is in the on-state.
Minimum On-State Voltage The minimum positive principal voltage
for which the differential resistance is
zero with the gate open-circuited.
Principal Current
A generic term for the current through
the collector junction.
NOTE: It. is the current through both
main terminals.
On-State Current
The pFincipal. current when the thyristor
is in the on-state.
Forward Current (of.a
The principal current for a positive anodeReverse Blocking Thyristor) to-cathode voltage.
Peak Repetitive On-State
The peak value of the on-state current including all repetitive transient currents.
Current
An on-state current of short-time duration
Surge (Non-Repetitive)
and specified waveshape.
On"State Current
Critical Rate of Rise of
The maximum value of the rate of rise of
On-State Current
on-state current which a thyristor can
withstand without deleterious effect.
Off-State Current
The principal current when the thyristor
is in the off-state.
Breakover Current
The principal current at the breakover
point.
Holding Current
The minimum principal current required
to maintain the thyristor in the on-state.
Latching Current
The minimum principal current required
to maintain the· thyristor in the on-state
immediately after switching from the offstate to the on-state has occurred and the
triggering signal has been removed.
Reverse Voltage (of a
A negative anode-to-cathode voltage.
Reverse Blocking Thyristor)
The maximum instantaneous value of the
Working Peak Reverse
reverse voltage which occurs across the
Voltage (of a Reverse
Blocking Thyristor)
thyristor, excluding all repetitive and
non-repetitive transient voltages.
The maximum instantaneous value of the
Repetitive Peak Reverse
reverse voltage which occurs across the
Voltage (of a Reverse
thyristor, including all. repetitive transient
Blocking Thyristor)
voltages, but excluding all non-repetitive
transient voltages.
30
SYMBOLS AND TERMINOLOGY
Non-Repetitive Peak
Reverse Voltage (of a
Reverse Blocking Thyristor)
Reverse Breakdown Voltage
(of a Reverse Blocking
Thyristor)
The maximum instantaneous value of any
non-repetitive transient reverse voltage
which occurs across a thyristor.
The value of negative anode-to-cathode
voltage at which the differential resistance
between the anode and cathode terminals
changes from a high value to a substantially lower value.
Reverse Current (of a
The current for negative anode-to-cathReverse Blocking Thyristor) ode voltage. .
Reverse Breakdown Current The principal current at the reverse
(of a Reverse Blocking
breakdown voltage.
Thyristor)
Gate Voltage
The voltage between a gate terminal and
a specified main terminal.
NOTE: Gate voltage polarity is referenced to the specified main terminal.
Gate Current
The current that results from the gate
voltage.
NOTE: 1. Positive gate current refers
to conventional current entering the gate terminal.
2. Negative gate current refers
to conventional current leaving the gate terminal.
Gate Trigger Voltage
The gate voltage required to produce the
gate trigger current.
Gate Non-Trigger Voltage
The maximum gate voltage which will
not cause the thyristor to switch from the
off-state to the on-state.
Gate Trigger Current
The minimum gate current required to
switch a thyristor from the off-state to
the on-state.
Gate Non-Trigger Current
The maximum gate current which will
not cause the thyristor to switch from the
off-state to the on-state.
Thermal Resistance (of a
The temperature difference between two
specified points or regions divided by the
Semiconductor Device)
power dissipation under conditions of
thermal equilibrium.
Transient Thermal Impedance (of a Semiconductor
Device)
The change of temperature difference between two specified points or regions at
the end of a time interval divided by the
step function change in power dissipation
at the beginning of the same time interval
causing the change of temperature difference.
31
SCR MANUAL
Gat~ Controlled Turn-On
Time
Gate Controlled Delay
Time
Gate Controlled Rise Time
Circuit-Commutated
Tum-Off Time
Reverse Recovery Time
(of a Reverse Blocking
Thyristor)
I squared t (I 2t)
Mounting Force
Stud Torque
32
The time interval between a specified
point at the beginning of the gate pulse
and the instant when the principal voltage (current) has dropped (risen) to a
specified low (high) value during switching of a thyristor from off-state to the
on-state by a gate pulse.
The time interval between a specified
point at the beginning of the gate pulse
and the instant when the principal voltage (current) has dropped (risen) to a
specified value near its initial value during switching of a thyristor from the offstate to the on-state by a gate pulse.
The time interval between the instants at
which the principal voltage (current) has
dropped (risen) from a specified value
near its initial value to a specified low
(high) value during switching of a thyristor from the off-state to the on-state by
a gate pulse.
NOTE: This time interval will be equal
to the rise time of the on-state
current only for pure resistive
loads.
The time interval between the instant
when the principal current has decreased
to zero after external switching of the
principal voltage circuit, and the. instant
when the thyristor is capable of supporting a specified principal voltage without
turning on.
The time required for the principal current or voltage to recover to a specified
value after instantaneous switching from
an on-state to a reverse voltage or current.
This is a measure of maximum forward
non-recurring overcurrent capability for
very short pulse durations. The value is
valid only for the pulse duration specified.
I is in RMS amperes, and t is pulse duration in seconds. (I 2t is necessary for fuse
co~ordination. )
Range of mounting forces. recommended
for Press Pak packages to insure an adequate thermal and electrical path while
avoiding mechanical damage.
Recommended mounting torque for stud
packages.
SYMBOLS AND TERMINOLOGY
2.2.3 Letter Symbol Table
Quantity
On-State Current
DC Value, DC Value,
No Alter- With Alter- Instannating
nating
taneous
Total RMS CompoCompoTotal
Value
nent
nent
Value
IT(HMS)
IT
IT(AY)
iT
Maxi·
mum
(Peak)
Total
Value
In!
Repetitive Peak
On-State Cunent
I'nu[
Surge (Non-Repetitive)
On-State Current
Inm
Breakover Current
Off-State Current
I mo )
ID(RlIIs)
In
i (JIOl
ID(AY)
in
Repetitive Peak
Off-State Current
Reverse Current
lI>mI
IR(HMs)
In
In(AV)
iR
Repetitive Peak Reverse
Cunent
Reverse Breakdown
Current
On-State Voltage
:areakover Voltage
Off-State Voltage
Minimum On-State
Voltage
II»[
IID[
Imor
i(HRIIt
lemon
VT(mIS) VT
V mo )
VD(RMS) VI)
VT(AY)
VT
V'DI
vmo)
VII(AY) VI)
V n )[
VTOIHN)
Working Peak OffState Voltage
V IIW )!
Repetitive Peak OffState Voltage
V 1111)1
Non-Repetitive Peak
Off-State Voltage
V IIS )[
Reverse Voltage
VR(ItMS) V It
VIt(AY) VR
VIOl
Working Peak Reverse
Voltage
V mnl
Repetitive Peak Reverse
Voltage
V HIDI
33
SCR MANUAL
Quantity
DC Value, DC Value,
No Alter- With Alter- Instannating
nating
taneous
Total RMS CompoCompoTotal
Value
nent
Value
nent
Non-Repetitive Peak
Reverse Voltage
Maximum
(Peakl
Total
Value
VRs~r
Reverse Breakdown
Voltage
V(Bll)R
V(JlR)R
Holding Current
IH
iIi
Latching Current
IL
iL
Gate Current
IG
Gate Trigger Current
IG(AV)
iG
1m!
IGT
iGT
IGT~r
Gate Non-Trigger
Current
IGD
iGD
I GInr
Gate Voltage
VG
VG
Vm!
Gate Trigger Voltage
V GT
VGT
VGTM
Gate Non-Trigger
Voltage
V GD
VGD
V Gmr
Gate Power Dissipatior
PG
PG
PG)f
VG(AV)
PG(AV)
2.2.4 General Letter Symbols
Present Symbol
Ambient Temperature
Case Temperature
Junction Temperature
Storage Temperature
Thermal Resistance
Thermal Resistance, Junction-to-Case
Thermal Resistance, Junction-to-Ambient
Thermal Resistance, Case-to-Ambient
Transient Thermal Impedance
Transient Thermal Impedance,
J unction-to-Case
Transient Thermal Impedance,
Junction-to-Ambient
Delay Time
Rise Time
Fall Time
Reverse Recovery Time
Circuit-Commutated Turn-Off Time
34
Former Symbol
TA
Tc
TJ
T stg
Re
ReJC
ReJA
RecA
8
8 J -c
8 J -A
8 C-A
Zeit)
8(t)
ZeJc(t)
B.l-A(t)
tr
tf
trr
tq
RATINGS AND CHARACTERISTICS OF THYRISTORS
3
RATINGS AND CHARACTERISTICS OF THYRISTORS
The family of thyristor devices has in common a switching capability in one or two quadrants of its V-I characteristics. Thyristor
devices used as power switches have in common the necessity for
proper design and specification of their heat dissipation and heat transfer properties. Furthermore, thyristors are switched into the on-state
either by applying a triggering signal to their gate or by increasing
off-state voltage until it exceeds the breakover voltage characteristic.
These and other common properties of thyristor devices allow a uniform
approach to thyristor characterization which need differ only in detail
when applied to a specific thyristor device like, for example, an SCR
or a triac.
In the following sections of this chapter the discussion is largely
in terms of SCR's. Most of this material is applicable, however, to
other thyristor devices. Specialized characterization information is presented in Chapter 7 for triacs and in Chapter 14 for light-activated
thyristors.
3.1 JUNCTION TEMPERATURE
The operating junction temperature range of thyristors varies for
the individual types. A low temperature limit may be required to limit
thermal stress in the silicon crystal to safe values. This type of stress is
due to the difference in the thermal coefficients of expansion of the
materials used in fabricating the cell subassembly. The upper operating temperature limit is imposed because of the temperature dependence of the break over voltage, turn-off time and thermal stability
considerations. The upper storage temperature limit in some cases may
be higher than the operating limit. It is selected to achieve optimum
reliability and stability of cparacteristics with time.
/
The rated maximum operating junction temperature can be used
to determine steady-state and recurrent overload capability for a given
heatsink system and maximum ambient temperature. Conversely, the
required heatsink system may be determined for a given loading of
the semiconductor device by means of the classic thermal impedance
approach presented in Sections 3.3 and 3.4.
Transiently the device may actually operate beyond its specified
maximum operating junction temperature and still be applied within
its ratings. An example of this type of operation occurs within the
specified forward non-recurrent surge current rating. Another example
is the local temperature rise of the junction due to the switching dissipation during the turn-on of a thyristor under some conditions. It is
impractical at this time to establish temperature limits for these types
35
SCR MANUAL
of operating stresses from both a. rating as well as an applications point·
of view. Therefore, such higher-than-rated temperature operation must
remain implicit in other ratings established for the device.
3.2 POWER DISSIPATION
The power generated in the junction region in typical thyristor
operation consists of the following five components of dissipation:
a. Tum-on switching
h. Conduction
c. Turn-off or Commutation
d. Blocking
e. Triggering
On-state conduction losses are the major source of junction heating for normal duty cycles and power frequencies. However, for very
steep (high di/dt) current waveforms or high operating frequencies
turn-on switching losses may become the limiting consideration. Such
cases are discussed in Sections 3.7 and 3.8.
AVERAGE ON-STATE CURRENT- AMPERES
FIGURE 3.1
MAXIMUM AVERAGE ON·STATE POWER DISSIPATION FOR C137 SERIES SCR
Figure 3.1 gives on-state conduction loss in average watts for the
C137 SCR as a function of average current in amperes for various conduction angles for operation up to 400 Hz. This type of information is
given on the specification sheet for each type of SCR (with the exception of some inverter type SCR's). These curves are based on a current
waveform which is the remainder of a half-sine wave which results
when delayed angle triggering is used in a single phase resistive load
circuit. Similar curves exist for rectangular current waveforms. These
power curves are the integrated product of the instantaneous anode
current and on-state voltage drop. This integration can be performed
graphically or analytically for conduction angles other than those listed,
36
RATINGS AND CHARACTERISTICS OF THYRISTORS
using the on-state voltage-current characteristic curves for the specific
device.
Both the on-state and reverse blocking losses are determined by
integration of the appropriate blocking E-I curves on the specification
sheet.
Gate losses are negligible for pulse types of triggering signals.
Losses may become more significant for gate signals with a high duty
cycle, or for SCR's in a small package such as the TO-5, TO-IS or
Power Tab type packages. The losses may be calculated from the
gate E-I curves shown on the triggering characteristics for the specific
type of SCR. Highest gate dissipation will occur for an SCR whose gate
characteristics intersect the gate circuit load line at its midpoint. For a
more detailed discussion of the gate characteristic and its load line, see
Chapter 4.
Turn-on switching ratings are discussed in Section 3.S. Turn-off
is discussed in Chapter 5.
3.3 THERMAL RESISTANCE
The heat developed at the junctions by the foregoing power losses
Hows into the case, then to the heatsink (if employed) and on to the
surrounding ambient Huid. The junction temperature rises above the
stud, or case, temperature in direct proportion to the amount of heat
Howing from the junction and the thermal resistance of the device to
the How of heat. The following equation defines the relatio-'clship under
steady-state conditions:
= PRaJC
where
T J - Tc
TJ =
Tc
P=
Ra.Jc =
=
(3.1)
average junction temperature, °C
case temperature, °C
average heat generation at junction, watts
steady-state thermal resistance between junctions
and bottom face of hex or case, °C/watt
Equation 3.1 can be used to determine the allowable power dissipation and thus the continuous pure DC on-state current rating of an
SCR for a given case temperature through use of the on-state E-I
curves. For this purpose, T J is the maximum allowable junction temperature for the specific device. The maximum values of RaJc and T J
are given in the specifications.
3.4 TRANSIENT THERMAL IMPEDANCE
3.4.1 Introduction
Equation 3.1 is not satisfactory for finding the peak junction temperature when the heat is applied in pulses such as the recurrent conduction periods in an AC circuit. Solution of Equation 3.1 using the
peak value of P is over-conservative in limiting the junction temperature
rise. On the other hand, using the average value of P over a full cycle
will underestimate the peak temperature of the junction. The reason
for this discrepancy lies in the thermal capacity of the semiconductor,
37
SCR
MANUAL
that is, its characteristic of requiring time to heat up, its ability to
store heat, and its cooling before the next pulse. .
Compared to other electrical components such as transformers
and motors, semiconductors have a relatively low thermal capacity,
particularly in the immediate vicinity of the junction. As .a result,
devices like the SCR heat up very quickly upon application of load,
and the temperature of the junction may fluctuate during the course
of a cycle of power frequency. Yet, for very short overloads this relatively low thermal capacity may be significant in arresting the rapid
rise of junction temperature. In addition, the heatsink to which the
semiconductor is attached may have a thermal constant of many minutes. Both of these effects can be used to good advantage in securing
attractive intermittent and pulse ratings sometimes well in excess of the
published continuous DC ratings for a device.
3.4.2 The Transient Thermal Impedance Curve
The thermal circuit of the SCR can be simplified to that shown in
Figure 3.2. This is an equivalent network emanating in one direction
from the junctions and with the total heat losses being introduced at
the junctions only. This simplification is valid for current amplitudes
at which I2R losses are small in comparison with the junction losses.
In Figure 3.2 the case of the power semiconductor is the reference
level. If a small stud type device is mounted to an infinite heatsink, the
heatsink temperature can be used as a reference. However, with larger
devices, the case to heatsink thermal resistance is relatively large compared to the junction-case thermal resistance. In such cases the case
or hex temperature should be used as a reference.
When a step pulse of heating power P is introduced at the junctions of the SCR (and of the thermal circuit) as shown in Figure 3.3A,
the junction temperature will rise at a rate dependent upon the response
of the thermal network. This is represented by the curve T heat in Figure 3.3B. After some suffiCiently long time tb the junction temperature
will stabilize at a point aT
PReJc above the ambient (or case) tem~
perature. This is the steady-state value which is given by Equation
3.1. ReJc is the sum of,Re1 through ReN in the equivalent thermal circuit of Figure 3.2. Chapter 20 gives specific instructions complete with
circuit schematics for measuring SCR characteristics.
=
FIGURE 3.2 SIMPLIFIED EQUIVALENT THERMAL CIRCUIT FOR A POWER SEMICONDUCTOR
38
RATINGS AND CHARACTERISTICS OF THYRISTORS
If the power input is terminated at time t2 after the junction temperature has stabilized, the junction temperature will return to ambient
along the locus indicated by Tcool in Figure 3.3B. It can be shown that
curves T heat and Trool are conjugates of one another,! that is,
(3.2)
By dividing the instantaneous temperature rise of curve T heat in
Figure 3.3B by the power P causing the rise, the dimensions of the ordinate can be converted from °C to DC/watt. This latter set of dimensions is that of thermal resistance, or as it is more precisely termed: the
transient thermal impedance ZS(t). Figure 3.4 shows a plot of transient
thermal impedance for the C34 SCR both when mounted to an infinite
heatsink and to a four-inch square copper fin.
Transient thermal impedance information for a device can be
obtained by monitoring junction temperature at the end of a welldefined power pulse or after a known steady-state load has been
removed. Junction temperature is measured by use of one of the temperature-sensitive junction characteristics such as on-state voltage drop
at low currents. Conversion of heating data to cooling data, or vice
versa, can be accomplished through the use of Equation 3.2.
0
HEAT
INPUT
(WATTS)
0
'.
'.
I
I
TIME-
I
I
I
I
I
,I
®
JUNCTION
TEMPERATURE
("C)
-AMBIENT
'0
FIGURE 3.3
" '2
TIME-
RESPONSE FOR SCR JUNCTION TO STEP PULSE OF HEATING POWER
In order that the transient thermal impedance curve may be used
with confidence in equipment designs, the curve must represent the
highest values of thermal impedance for each time interval that can be
expected from the manufacturing distribution of the products.
The transient thermal impedance curve approaches asymptotic
values at both the long time and short time extremes. For very long
time intervals the transient thermal impedance approaches the steadystate thermal resistance R sJc.
39
SCR MANUAL
10
8
STUD MOUNTED
ON 4- X 4" X 1/16"
PAINTED COPPER FIN
(JUNCTION TO AMBIENT)
...
..-~
~
~
I
Q.
0.8
W
u
I
i-""
I
O.G
"""
V
"
.""
,
STARTING FROM CASE TEMP.• EQUALS loooe (MAX. Te) MINUS'
CASE TEMP DIVIDED BY THE TRANSIENT THERMAL
IMPEDANCE.
loooe -Te
PpEAK = R9J _C {t l
0.04
- c-
00 2
0.001
'"''
~~!~ ~~~~~~~TL~REOI~~~P!~~~EINL~~~:'~~~~.:R~~:~~~ 1.1
-CCf=
0.0 I
"
NOTES . I. CURVE DEFINES TEMPERATURE RISE OF JUNCTION ABOVE
2
.I
CELL MOUNTED ON INFINITE
HEAT SINK (CASE TEMPERATURE
FIXED) (JUNCTION TO CASE)
2. FOR OPTIMUM RATINGS AND FURTHER INFORMATION,SEE
PUB. #200.9 ENTITLED "POWER SEMICONOUCTOR RATINGS
UNDER TRANSIENT AND INTERMITTENT LOADS:'
1111111
0.004
0.04
O.QI
0.1
I
II111111
0.4
I
11111111
I III11I
10
40
100
400
1000
TIME (f) -SECONDS
FIGURE 3.4
MAXIMUM TRANSIENT THERMAL IMPEDANCE OF C34 SCR
For times less than 1 millisecond the value at 1 millisecond may
be extrapolated by 1/ V t . This is based on assuming that the time of
interest is sufficiently small so that all heat generated at the semiconductor junction may be considered to flow by one-dimensional diffusion
within the silicon pellet. This assumption is valid for times small compared to the thermal time constant of the path junction-to-face of the
silicon pellet. For extremely short times, however, during which current
density, and hence heating, is non-uniform, this approach is no longer
valid.
For example the C34 transient thermal impedance at 40 p.Seconds
may be estimated at
.083
1 X 10- 3 = .00OO°C/Watt
4 X 10'-5
However, the extrapolated values are valid only for times after the SCR
has turned on fully. These values should, in other words,not be used
during the switching interval (see Section 3.8). In general the switching
interval is between 20 and 200 p.seconds for medium and high current
SCR's respectively and a minimum of 10 pSeconds for the very low
current devices. The switching interval can be estimated by inspecting
the energy per pulse data curves for the knee of the· constant wattseconds curves. For example, Figure 3.5 shows theknee to be at about
20 pSeconds for the C141 SCR. Since the e34 has a similar current
rating extrapolations down to 20 p.seconds would seem· to be in order.
However due to differences in gate geometries 40 ,...seconds is a conservative lower limit. (See Chapter 1, Section 1.5 for further data on
gate geometries.)
40
RATINGS AND CHARACTERISTICS OF THYRISTORS
I
I I II
~O~~~LJE 111,wATLEc lJJt XLRAt
YIN
(2) MAX. ALLOWABLE CASE TO AMBIENT THERMAL
RESISTANCE" 5·C PER WATT WITH RATED
BLOCKI~'G VOLTAGE APPLIED
(3) SINUSOIDAL WAVE SHAPE
1000
.................. i""-..~II
. . . . . . . . . . . ""i'......~,.-4,.,.
400
-~r---.~ ~ ,,~~ "~.p
--
--
100
40
10
I
--------
4
FIGURE 3.5
<~~~,~,~~~
~O 0"
~O
""'"
~},
",0eo..
"- ,~ ~~,~ ~ l0
0"
""'"
'0
10
~~
40
"""
""
"" "'~~ "" ~~~ "
100
PULSE
""
400
1000
4000
10,000
~
BASE WIDTH - MICROSECONDS
ENERGY PER PULSE FOR SINUSOIDAL PULSES FOR GE C141 SCR
For maximum utilization of semiconductor devices in the switching region additionalfactors must be considered and other methods of
rating and life testing must be used. Consult Sections 3.7 and 3.8 for
further information.
3.4.3 The Effect of Heatsink Design on the Transient
Thermal Resistance Curve
Since the heatsink is a major component in the heat transfer path
between junction and ambient, its .design affects the transient thermal
impedance curve (Figure 3.4) substantially. When a semiconductor is
manufactured and shipped to the user the. manufacturer has no control
over the ultimate heatsink and can only provide data on the' heat transfer system between junction and case which, is the part he manufactured. These type of data are presented in Figure 3.4 as the "Cell
Mounted to Infinite Heatsink" curve.
The equipment designer can use this curve in developing a transient thermal impedance curve for the cell when mounted to a particular heatsink of his own design by means of a few simple calculations.
These calculations consist of first determining the heats ink time constant by deriving the relationship as shown in Chapter 18 (Mounting
and Cooling The Power Semiconductor) and then adding the transient
thermal relationship of the heatsink to that of the cell.
For example: From Chapter 18 we find that a 4" x 4" painted
copper fin Yt6" thick has a transient thermal resistance given by
Ze(t)fln = 3.1 (I - e-t/ l74)
Assume case to fin contact resistance is negligible for example simpli~
fication.
41
SCR MANUAL
The values of Ze(t)fln are added to the infinite heatsink curve to
secure the over-all Ze(t) of the system as indicated in Figure 3.4. Note
that this fin makes a negligible contribution to the over"all thermal
impedance of the cell-heatsink system at periods of time one second
or less after application of power. In this area the fin behaves like an
infinite heatsink, that is, one of zero thermal resistance. The fin-heatsink
system reaches equilibrium around 1000 seconds. Thereafter the
thermal capacity is no longer effective in holding down the junction
temperature.
3.5 RECURRENT AND NON·RECURRENT CURRENT RATINGS
3.5.1 Introduction
The discussion under all parts of this section and Section 3.6
applies to the .conventional rating system presently used for thyristors
when turn-on switching dissipation is negligible. Turn-on switching
characterization is discussed in Section 3.7; current ratings for high
frequency operation which cannot neglect turn-on switching losses are
discussed in Section 3.8.
When a semiconductor device is applied in such a manner that
its maximum allowable peak junction temperature is not exceeded the
device is applied on a recurrent basis. Any condition that is a normal
and repeated part of the application or equipment in which the semiconductor device is used must meet this condition if the device is to
be applied on a recurrent basis. Section 3.6 gives methods of checking
peak junction temperature. These enable the designer to properly apply
the device on a recurrent duty basis.
A class of ratings that makes the SCR and the triac truly power
semiconductors are the non-recurrent current ratings. These. ratings
allow the maximnm (r('('urr('nt) op('rating jUl1etiOIl temperature of Ihe
dcvicc to be exceeded for a prief instant. This gives the device an
instantaneous overcurrent capability allowing it to be coordinated with
circuit protective devices such as circuit breakers, fuses, 2 etc. The
specification bulletin gives these ratings in terms of surge current and
Pt. These ratings, then, should only be used to accommodate unusual
circuit conditions not normally a part of the application, such as fault
currents. Non-recurrent ratings are understood to apply to load conditions that will not occur more than a limited number of times in the
course of the operating life of the equipment in which the SCR is
finding application. (JEDEC*defines the number of times as equal to
or exceeding 100 times.) Also, non-recurrent ratings are understood to
apply only when they are not repeated before the peak junction temperature has returned to its maximum rated value or less. THE
LENGTH OF THE INTERVAL BETWEEN SURGES DOES NOT
CHANGE THE RATING. For example, if a garage door opener subjects a semiconductor device to its non-recurrent current rating, this is
misapplying the device. As a result, the device may be subject to failure
after 100 operations even though the device operates but once per
8 hour period.
• (JEDEC-Joint Electron Device Engineering Council-Semiconductor Standards
Organi7.ation )
42
RATINGS AND CHARACTERISTICS OF THYRISTORS
3.5.2 Average Current Rating (Recurrent)
Average current rating versus case temperature as it appears in
the specification sheet as for the C380 series SCR is shown in Figure
3.6. These curves specify the maximum allowable average anode current ratings of the SCR as a function of case temperature and conduction angle. Points on these curves are selected so that the j1lnction
temperature under the stated conditions does not exceed the maximum
allowable value. The maximum rated junction temperature of the C380
SCR is 125°C.
The curves of Figure 3.6 include the effects of the small contribution to total dissipation by reverse blocking, gate drive, and switching
up to 400 Hz. For devices which are lead mounted or housed in small
packages, like the TO-5 or Power Tab, the on-state current rating may
be substantially affected by gate drive dissipation. Where this becomes
important it is so indicated on the specification sheet.
The slope of the curves shown in Figure 3.6 is essentially dependent upon the ReJC . P D product. In some SCR's such as Press Paks and
Power Tabs, ReJc is not fixed but is a function of the method used to
cool it. Another family of curves would be needed in place of Figure
3.6 for single side cooling of the Press Pak package. Similarly, small
packages such as the Power Tab may have more than one set of curves
to take into account different mounting configurations and their corresponding effect on ReJC. 3
140 , - - - . - - - - , - - _ , _ _ - - , - - - , - - - - , - - - , - - - - - ,
~ 120~~~~~r_-~-~r_--r--r_-~-~
"'
0:
::J
....
~ 100r--\r~~~~~-~~-~--r_-_r-~
"'
:;
"'....
"'
«
80~-~~-~-~~~~~-L-~~~~-~
~
60~~~~~~-_,__-~
Q.
Cf)
u
m
~
o
j
«
40~-_r--~-_r--r
:;
i
X
«
20
:;
SINUSOI DAL
CURRENT
WAVEFORM
oL-___ L_ _ _ _L -___ L_ _ _ _L -___ L_ _ _ _ _ __ L_ _
o
50
100
150
200
250
300
350
~
~
400
AVERAGE ON-STATE CURRENT - AMPERES
FIGURE 3.6
MAXIMUM AVERAGE CURRENT RATINGS FOR C380 SERIES SCR
43
SCR MANUAL
If the C380 in a single phase resistive load circuit is triggered as
soon as its anode swings positive, the device will conduct for 180
electrical degrees. If the case temperature is maintained at 80°C, or
less, the C380 is capable of handling 235 amps average current as indicated· in Figure 3.6. If the triggering angle is retarded by 120° the
C380 will conduct for only the 60 remaining degrees of the half cycle.
Under these conditions of 60° conduction, the maximum rated average
current at 80°C stud temperature (double side cooled) is 115 amperes,
substantially less than for 180° conduction angle. This leads us nicely
into the next section.
3.5.3 RMS Current (Recurrent)
It will be noted in Figure 3.6 that the curves for the various conduction waveshapes have definite end points. These points represent
identical RMS values, and as such an RMS rating is implicit in the
curves of Figure 3.6.
For example, the C380 is rated 370 amperes DC or
;.~~
= 235
amperes average in a half-wave, or 180° conduction angle, circuit. The
factor 1.57 is the form factor giving the ratio of RMS to average values
for a half wave sinusoidal waveform. By the definition of RMS values,
the RMS and average values are identical for a direct current. The
RMS current rating, as shown on the specification sheet for individual
SCR's, is necessary to prevent excessive heating in resistive elements
of the SCR, such as joints, leads, interfaces, etc.
The RMS current rating can be of importance when applying
thyristors to high peak current, low duty cycle waveforms. Although
the average value of the waveform may be well within the ratings,
it may be that the allowable RMS rating is being exceeded.
The average current values shown as phase control ratings in Figure 3.6 for a given and fixed basic RMS device current rating are for
the resistive current waveform shown in the figure. Since the current
form factor for the case of resistive loading is greatest, and since inductance in the path of the thyristor current will reduce its form factor, the
average current ratings in Figure 3.6 are conservative for inductive
current waveforms.
For inductive waveforms in which the thyristor current waveform
is essentially rectangular, such as may occur in a phase-controlled rectifier operating near full output, most specification sheets show separate
rating curves to reflect the improvement in form factor. However, such
current waveforms are, of course, subject to the restriction of the allowable tum-on current rating of the device.
In other cases in which the current waveform may be halfsinusoidal in shape but of a base width less than half a period of the
supply frequency as, for example, with discontinuous AC line current
in an AC switch application4 at large phase retard, greater utilization
of the thyristor in terms of its average current versus temperature
ratings (like in Figure 3.6) can be obtained by taking into account the
improvement of form factor due to decreasing load power factor
(greater inductance) when applying the device within its RMS current
rating. See also Section 9.2.1.
44
RATINGS AND CHARACTERISTICS OF THYRISTORS
3.5.4 Arbitrary Current Waveshapes and Overloads (Recurrent)
Recurrent application of arbitrary waveshapes, varying duty
cycles, and overloads requires that the maximum peak allowable junction temperature of the SCR not be exceeded. Section 3.6 gives information for determining this.
3.5.5 Surge and 12t Ratings (Non-Recurrent)
In the event that a type of overload or short circuit can be classified as non-recurrent, the rated junction temperature can be exceeded
for a brief instant, thereby allowing additional overcurrent rating.
Ratings for this type of non-recurrent duty are given by the Surge
Current and I 2t rating curves.
.
Figure 3.7 shows the maximum allowable non-recurrent multicycle surge current at rated load conditions. Note that the junction
temperature is assumed to be at its maximum rated value (125°C for
the C398); it is therefore apparent that the junction temperature will
exceed its rated value for a short time during and immediately following
operation within the non-recurrent ratings. Therefore many of the
SCR's ratings and characteristics will not be valid until the junction
tempera ture cools back down to within its rated value. The reader
is thus reminded that off-state blocking capability, dv/dt and turn-off
time, to name just a few device parameters, are not specified or
guaranteed immediately following device operation in the non-recurrent current mode.
The data shown by the solid curve "A" are values of peak rectified
sinusoidal waveforms on a 60 Hz basis in a half-wave circuit. The "onecycle" point, therefore, gives an allowable non-recurrent half sine wave
of 0.00834 seconds' duration (half period of 60 Hz frequency) of a peak
amplitude of 7,300 amperes. The "20 cycle" point shows that 20 rectified half sine waves are permissible (separated by equal "off" times),
each of an equal amplitude of 5,100 amperes.
The data shown by the dotted curve "B" for 50 Hz operation has
been added to the curve and is not regularly part of the published
data sheet. The curve is constructed by connecting two points on the
curve with a straight line. The current value for the first point at 1 cycle
is obtained from Figure 3.8 at 10 ms, the base width of a 50 Hz sine
wave. The second point coincides with the 60 Hz, one second (60
cycles) value. Beyond the one second value the curves for 50 and 60 Hz
waveforms are the same and are extensions of the 60 Hz curve.
45
SCR MANUAL
NOTE:
JUNCTION TEMPERATURE IMMEDIATELY
PRIOR TO SURGE =-40" TO + 125·C.
OLI~~2--~~~6~8~10~~~~-4~0-L6~0~8~OUIOO
CURVE A, CYCLES AT 60 Hz
CURVE B, CYCLES AT 50 Hz
FIGURE 3.7
MAXIMUM ALLOWABLE MULTICYCLE, NON-RECURRENT, PEAK SURGE ONoSTATE
CURRENT FOR THE C398 SERIES SCR
The lower half of Figure 3.8 shows the maximum allowable nonrecurrent sub-cycle surge current at rated load conditions. Like its
sister multicycle curve of Figure 3.7, it is apparent that the junction
temperature will again exceed its rated value for a short time.
300,00 0
250,000
1--''''''
0200,00 0
-"''"
~NQ.. 150.00 0
~OR
V
:Ii
5
OF
./
100,000
V
~
HALF SINE WAVE
IURR~NTI
T
80,00 0
I I
I
;:.
"';;;
..J'"
<.>'"
>-'"
<.>0.
15,000
,:Ii
CD
0",
><'"
u
ui
20
1200
50
2i
100
Ii
2i
:::>
2i
BOO
x
,N.(
_
~"~{O~
.
r-.
'~
~~» "~
,'o Q" ~ \- ,"-.;"
~,
..............
I
.......
NOTESI. SWITCHING VOLTAGE
lin
"\
,"",
.............. '-
................
,
I'-..
.........
:--....
r-.. . ,.o~s
o
20
--.....
=800 VOLTS _-:-- ---' _
2. MIN. CKT. TURN-OFF-TlME:; 40~SEC
3. MAX, CKT. OVlOT:: 200 VOLTS/JJ.SEC
4. REVERSE VOLTAGE APPLIED-50 V~~800V
~~~~""",,
-"":'£$e:' " "
'~
'-
'" ."""I'-.
" '" "
5. REQUIRED GATE C;>RIVE'
I'-.
I
SOURCE:20 VOLTS, 65 OHMS
CURRENT RISE TIME" I LISEe
6.Re SNUBBER CKT. :::.25~f.5n.
10
10
20
40
60
80
100
200
400
600 800 1000
2000
4000 6000 8000 10,000
PULSE BASE WIDTH - MICROSECONDS
FIGURE 3.20
ENERGY PER PULSE FOR SINUSOIDAL PULSES FOR THE GE C158/C159 SCR
3.8.2 High Frequency Rectangular Waveshape Current Ratings
Rectangular lO current wavefonns are the mainstay of switching
SCR's operating in low to medium frequency power conversion systems.
Popular examples of circuitry imposing this type of duty on the main
power switches are pulse width modulated inverters for AC motor
speed control and DC choppers for DC motor speed control.
To fully characterize an SCR under rectangular current wavefonn
conditions, four parameters are needed to define the operating wavefonn as shown in Figure 3.21 and a fifth, case temperature, is needed
to specify thermal conditions.
58
RATINGS AND CHARACTERISTICS OF THYRISTORS
PARAMETERS NEEDED:
ITM
PEAK CURRENT.
DIIDT
LEADING EDGE DI/DT.
liT
REP. RATE.
100( ~)
FIGURE 3.21
DUTY CYCLE.
RECTANGULAR CURRENT WAVEFORM DEFINITION
An example of a current rating curve for the C398 SCR is shown
in Figure 3.22. Additional curves are given in the data sheet for 25%
and 10% duty cycle operation to allow for interpolation between 75%
and 5% duty cycle operation. Like the sine wave rating curves, data
is also provided for other case temperatures to again allow for data
interpolation.
DUTY CYCLE - 50%
en...
......
NOTES'
OFF-STATE VO LTAGE =800 VOLTS
REVERSE VOLTAGE!:800VOLTS
DUTY CYCL E = 50 %
CASE TEMP. = 65°C
II::
1000
:IE
800
co:
"'-
'I
z
f-
600
II::
II::
500
...
400
...
::>
0
-
........
REQUIRED GATE DRIVE'
20 VOLTS,65 OHMS,
IILSEC RISETIME
RC SNUBBER'
.21LF,5 OHMS
60Hz
r ---
t-- "'-
....
400 Hz
F
I KHz
2.5 KHz
f-
I
300
~
(I)
5 KHz
I
i5
......""co:
200
100
4
5
6
8
10
15
20
30
40
60
80 100
RATE OF RISE OF ON-STATE CURRENT- (AMPERESIILSEC)
FIGURE 3.22
MAXIMUM ALLOWABLE PEAK ON-5TATE CURRENT VS di/dt (Te
= 65°C)
Switching loss data for heatsink selection is given in the form of
watt-seconds/pulse data as shown in Figure 3.23. Because of the additional parameter, dildt, needed to characterize the rectangular waveform, three such charts are needed where a single chart was adequate
for the sinusoidal waveform case. The two additional charts characterize the losses for 25 and 5 amps/p.Second respectively; again,
interpolation is employed to determine losses for dildes in between
those given.
59
SCR MANUAL
3,000
~
2Poo
0;
":Ii'"'"
Q.
~
,
....
~~
..-
~~
1,500
·z
700
'"0:
600
""
0
If
200
'"
Q.
t--.....
.........
......
r-
r-
.......
......
6, RC SNUBBER. 2,._, 5 OHMS
40 50 60
I
11II1
80 100
150
200
......
I
I
300 400
~
~1
600
~,
K
~
800 IPoo
PULSE BASE WIDTH-(,.SEC)
FIGURE 3.23
PULSE
,,'0
t'-.
r" " I't'-.
i'
~
~
"-
........
.......... 5
..... ~
4. dy/dl = 200 VI,.SEC
I I
"
""
~.,.T-SECI
1.0
5, %"J~T?:~VE =20 VOLTS,65 OHMS,I,.SEC
100
r-.r-.
2.5
NOTES:
I. AVG. POWER=WATT - S E C / "
PULSE X REP RATE .
300 r- 2. SWITCHING VOLTAGE=SOO VOLTS
3. tq =C398-40,.SEC, C397-60,.SEC
400
'"\(....
v:z
.....
1"'- ~ ~
r-...
.......
1-1-
500
u
l- t--...
-r--
I POD
900
800
r--
"-
~
~
'\
1'\
"
2,000
i'
r'\.
"-~
\
4,000
~~
6,000
10,000
8,000
ENERGY PER PULSE FOR RECTANGULAR PULSES FOR THE C397/C398 SCR
(di/ dt = 100 A/ !-,sec)
3.9 VOLTAGE RATINGS
The voltage ratings of SCR's have been traditionally designated
by a single suffix letter, or single letter group, in the model number
of the device (e.g., C35B) or are an integral part of its JEDEC registration, The designation is translated in the specifications and defines
the thyristor's rated peak voltage which the device will safely withstand in both the off-state and reverse directions without breaking
down. The off-state was formerly referred to as the forward direction,
i.e., anode positive with respect to the cathode of an SCR. It is applicable to any junction temperature within the specified operating range.
This symmetry of off-state and reverse voltage ratings is characteristic
of all standard low frequency SCR's. Symmetry does not always exist
for high frequency, inverter type SCR's. Where symmetry fails to
exist the device voltage grade may be specified by more than one
letter group separated by a number, dash or slash. An example is the
Cl38NIOM with a VDM of 600 volts and a VDRM of 800 volts. This particular device type also has a V RRM rating of 50 volts which is not
described in the type number designation.
Voltage ratings are related to'several device parameters and characteristics. Of primary concern is the blocking current and its relationship to' device junction temperature. Blocking cUQ'ent approximately
doubles with every lOoC rise in T J . Since junction temperature is a
direct function of total device power dissipation, it is possible to have
regenerative thermal runaway of an SCR if the SCR's heatsink is above
a critical value. l l Generally this value is many times higher than
typically used to dissipate the SCR's losses due to current conduction.
This is certainly true of all the low frequency, slow turn-off SCR's
currently made. For fast turn-off, high frequency operation blocking
60
RATINGS AND CHARACTERISTICS OF THYRISTORS
current is traded off against enhanced tum-off time performance. The
higher blocking losses that result require a special rating format for
such devices if full advantage is to be taken of the device's inherent
voltage capability. The following discussion first considers the standard
voltage ratings which are applicable to both low and high frequency
SCR's. Later the special requirements of some high frequency SCR's
are discussed.
3.9.1 Reverse Voltage (V RRM ) and (V RSM)
In the reverse direction (anode negative with respect to cathode),
the SCR behaves like a conventional rectifier diode. General Electric
assigns two types of reverse voltage ratings: repetitive peak reverse
voltage with gate open, VRRM (formerly designated by "VROM(rep)");
and non-repetitive peak reverse voltage with gate open, VRSM (formerly
designated "VR0l1(non-rep) ").
If these ratings are substantially exceeded, the device will go into
breakdown and may destroy itself. Where transient reverse voltages
are excessive, additional VRRM margin may be built into the circuit by
inserting a rectifier diode of equivalent current rating in series with the
controlled rectifier to assist it in handling reverse voltage. For a detailed
discussion on voltage transients, see Chapter 16; for series operation,
see Chapter 6.
3.9.2 Peak Off-State Blocking Voltage (V ORM) (Formerly Peak
Forward Blocking Voltage (V FXM))
The peak off-state blocking voltage VDRM is given on the specification bulletin at maximum allowable junction temperature (worst case)
with a specified gate bias condition. The larger SCR's are specified for
a peak off-state blocking voltage rating with the gate open; smaller
SCR's are usually characterized for a peak off-state blocking voltage
with a specified gate-to-cathode bias resistor. The SCR will remain
in the off-state if its peak off-state voltage rating is not exceeded.
3.9.3 Peak Positive Anode Voltage (PFV)*
An SCR can be turned on in the absence of gate drive by exceeding its off-state breakover voltage characteristic V (BO) at the prevailing
temperature conditions. Although SCR's, in contrast to diode thyristors,
are designed to be brought into conduction by means. of driving the
gate, breakover in the off-state direction is generally not damaging
provided the allowable di/dt under this condition is not exceeded (see
Section 3.7).
Some SCR's are assigned a PFV rating. This rating is usually at
or above the VDRM rating. Off-state voltage which causes the device
to switch from a voltage in excess of its PFV rating may cause occasional degradation or eventual failure. Figure 3.24 illustrates the relationship between PFV and peak off-state blocking voltage rating VDRM .
61
SCR MANUAL
SIGNIFICANCE OF VORM AND PFV
PFV
.I7llh..
LlNE-/JllJ.,.
//11IliA VOLTAGE /l11111A
Ol~
SCR WILL NOT
TURN ON UNLESS
GATE TRIGGERED.
PFV
VDRM
--~HT---------h~~--
o
SCR MAY TURN ON,
-BUT
NO DAMAGE TO
SCR IF di/dt
IS KEPT WITHIN
DEVICE CAPABILITY
o
SCR MAY TURN ON
-AND
SCR MAY BE DAMAGED.
PFV
FIGURE 3.24 SIGNIFICANCE OF VDRM AND PFV RATINGS
The PFV rating is often of practical importance when SCR's are
tested for their actual breakover voltage characteristic V (BO) at room
temperature; often a unit will have a V (BO) beyond its PFV rating at
temperatures lower than maximum rated junction temperature. A
proper test for V (BO) under these circumstances would be to conduct
it at elevated temperature provided that V (BO) is lower than PFV.
In applications where the PFV rating of an SCR may be exceeded
it is suggested that a network be connected anode to gate so that the
device will trigger by gate drive rather than by off-state breakover.
A zener diode may be used to effect gate triggering at a predetermined
level, or a Thyrector diode may be used to obtain a similar action.
*Previously referred to as peak forward voltage. PFV is used as an abbreviation.
3.9.4 Voltage Ratings for High Frequency, Blocking
Power Limited SeR's
Inverter circuits frequently impose short time repetitive peak offstate and reverse. voltages upon SCR's. These transients are often
induced by the forced commutation circuits. Typically these transients
62
RATINGS. AND CHARACTERISTICS OF THYRISTORS
are in the 5 to 100 microsecond range and occupy less than 33% of
the blocking interval. In some circuits feedback diodes placed across
the SCR limit the reverse blocking voltage to only a few volts, typically
2 volts and always less than 50 volts.
In order to allow high frequency SCR's to block these short time
repetitive transients and yet not arbitrarily limit their voltage rating
due to high blocking losses at the high voltage levels a new rating
definition has been introduced as shown in Figure 3.25.
REPETITIVE PEAl<
OFF- STATE V
. / VDRM
MAX. DC
800
SWITCHING
/~~6~G:o
o -+-------'---,---+----;5~OV
I
25
FIGURE 3.25
I
50
75
I
600V
100 _
_____
TIME
PERCENT OF CYCLE
10Hz TO 25KHz
ALLOWABLE VOLTAGE ENVELOPE FOR C13BN10M AND C139N10M SCR'S
Basically the difference between this rating and the conventional
is the limitation on the duty cycle of VDRM to confined limits. The VDRlI
value is specified by the first letter code of the C139. The second letter
code indicates the VDM value. Any voltage envelope may be applied
to the device providing it is held to within the envelope prescribed in
Figure 3.25 for the C139NlOM and within the same envelope with the
addition of the 50 volt reverse limit for the C138NIOM.
Furthermore, the C139NIOM case to ambient thermal resistance
must not exceed 3.0°C/watt. Should the designer choose to operate
outside the voltage envelope shown, the factory must be consulted and
a lower value of Re (case to ambient) may have to be used in order to
maintain device thermal equilibrium. As the state of the art advances
and as experience is obtained with this rating philosophy, it is expected
that additional information will be provided the designer to enable
direct calculation of both the blocking losses and the related maximum
Re (case to ambient) for maintenance of device thermally stability.
3.10 RATE OF RISE OF OFF·STATE VOLTAGE (dv/dt)
A high rate of rise of off-state (anode-to-cathode) voltage may
cause an SCR to switch into the "on" or low impedance conducting
state. In the interest of circuit reliability it is, therefore, of practical
importance to characterize the device with respect to its dv/dt withstand capability.
The circuit designer may often limit the maximum dvI dt applied
to an SCR by means of added suppressor or "snubber" networks placed
across a device's terminals. Chapter 16 includes useful design information for the design of such networks.
General Electric SCR's and triacs are characterized with respect
to dv/dt withstand capability in the follOWing contexts:
63
SCR MANUAL
3.10.1 Static dv/dt Capability
This specification covers the case of initially energizing the circuit
or operating the device from an anode voltage source which has superposed fast rise-time transients. Such transients may arise from the
operation of circuit switching devices or result from other SCR's operating in adjacent circuits. Interference and' interaction phenomena of
this type are discussed further in Chapter 17. The industry standard
dv/dt definitions are defined by the waveforms shown in Figure 3.26.
NUMERICAL VALUE
OF EXPONENTIAL
50"~
.
~~
,.of-------F---""o-=-'C::::O"=r---75f----M~~~~
0.9 - - - - - - - - -
,
,
I
Va = TEST VOLTAGE PEAK
,
I
NUMERfCA'L VALUE
I
I
"'
I
"g
DVIDT
I
=0.,8 Yo,
Z- I
I
I
'2
(a) Exponential Waveform Test
(b) Linear Waveform Test
. FIGURE 3.26 dv/dt WAVEFORM DEFINITION
Either a linear ramp or the exponential waveform may be used.
When the exponential ramp is used the slope is defined as shown by
the linear ramp of Figure 3.26(a) intersecting the single time constant
value as shown. The linear waveform definition of Figure 3.26(b) is
self explanatory. The following discussion applies to the exponential
case used for industry registration purposes.
Some specification sheets give the time constant 'T under specified
conditions rather than a numerical value for dv/dt.
It will be noted that
'T
=
0.632 X Rated SCR Voltage (Vo )
dv/dt
(3.5)
The initial dv/dt withstand capability will be recognized as being
greater than the value defined in Figure 3.26(a). In terms of specified
minimum time constant it is
dv I dt
I
dv/dt
I
= Rated SCR Voltage (V0)
(3.6)
t=O+
'T
In terms of specified maximum dv/dt capability, the allowable initial
dv/dt withstand capability is
1 dv/dt = 1.58 dv/dt
(3.7)
= 0632
t=O+ .
The shaded areas shown in Figure' 3.26 represent the area of
dv/dt values that will not· trigger the SCR. These data enable the
circuit designer to tailor his circuitry in such a manner that reliable
circuit operation is assured .
.64
RATINGS AND CHARACTERISTICS OF THYRISTORS
Static dv/dt capability is an inverse function of device junction
temperature as well as a complex function of the transient waveform
shape. Figure 3.27 shows one example of how wave shape can greatly
change the withstand capability of a typical SCR.
700 PS-CR-TYiiEC50-
1600
500 f.400 -
o
::l
300 -
TYPICAL dv/dt.
WITHSTAND
CAPABILITY
(TJ" 125" C)
,.--_.
f---
-
I-t1&1
:l
.....
> 200
I
>:::i
t-
ID
100
c[
~
BO
c
z
60
50
Iii
40
o
--
-~
t
I
!f~I
,-
VOLTAGE
STEP-
I-~
!j
t
0
BIAS
-> VOLTAGE
0
I
!
TIME ..
0
\
\.
'\.
\
\"
~
BIAS
30 f-VOLTAG
~
20
.....
-\ -
\ ~ r\ \
c[
i
dv/dt
OF
SjEP ....
+
.\
'\.
"'\.
"
"-
00V\+200V",""
...
>
I'-........
'.............
.............
.........
c. . . . . . .
.......
r---........
""'-200V
--r-
IL
1&1
ttl)
300
400
500
600
700
MAGNITUDE OF VOLTAGE STEP -VOLTS-
FIGURE 3.27
800
TYPICAL dv/dt WITHSTAND CAPABILITY OF C50 SCR
Reverse biasing of the gate with respect to the cathode may
increase dv/dt withstand capability beyond that shown on an SCR's
data sheet. This increase is generally limited to medium and low current SCR's. The reader is referred to Chapter 1 for further discussion.
3.10.2 Reapplied dv /dt
This specification generally forms part of an SCR's turn-off time
specification and is really a turn-off time condition, rather than a specification in its own right. It is defined as: the maximum allowable rate
of reapplication of off-state blocking voltage, while the SCR is regaining its rated off-state blocking voltage VDRM, following the device's
turn-off time tq under stated circuit and temperature conditions. The
waveform is defined in Figure 3.28. For further information consult
Chapter 5.
65
SCR MANUAL
If)
D~
'z"
I
IW
c
o:
t!f2
0:
0:
0
U
W
:;)
W
C
0
Z
'"
~~
W
0:
-
TIME
I
~
If)
!::i
0
>
I
w
C>
'"
!::i
g
w
C
0
Z
'"
Iq -------I
I
I
I
I
I
C
0:
t!
I
I
I
0
lL
VORM
-REAPPLIED
dv/dl
I
0
w
~~
W
0:
FIGURE 3.28
REAPPLIED dv/dt WAVEFORMS
3.10.3 Triac Commutating dv/dt
Commutating dv/dt 12 differs from both the static and reapplied
dvI dt in that it presupposes device commutation immediately prior to
application of off-state voltage as shown in Figure 3.29. Commutating
dv/dt is generally substantially below a triac's static dv/dt rating.
Commutating dil dt, case temperature and RMS on-state current are
all conditions for the commutating dvI dt specification.
I
I
---i---
--tTIME-
I
I
I
I
___+,In
I
I
I
PRINCIPAL
CURRENT
I
ilL
I
I
I
~ITRM
I
---
IU
I
I
------
---
I
I
CJMMUTATIN.l
d"dt
I
I
:
I
I
PRINCIPAL
I
VOLTAGE
V ORM - - - . " " " ' - - - -
FIGURE 3.29
66
WAVEFORM OF COMMUTATING dv/dt
RATINGS AND CHARACTERISTICS OF THYRISTORS
Since commutating dv/dt varies with commutating di/dt, the factory should be consulted for operation of triacs beyond 60 Hz. Standard
selections are available for 400 Hz operation upon request. Figure 3.30
shows the typical variation of triac commutating dv/dt with commutating dil dt. Consult Chapter 7 for additional detail.
~------i00
50
;
NOTES;
I.Tc·7!5-t
_
~ 20
2. FOR 3fi()t'CONDUCTION.dl/dl-
!:i
rTiRMS) "'WHERE
g
i;
10
dl/dt IS IN AMPERES -;
IMIWSECONO AND "
rTIRMS) IS IN
~
AMPERES
~
5
707
~
~
"
~
~
~
"'" 'r--.,"'I'--
u
L...------I
I
5
10
20
50
100
di/dt - AMPERES/MILLISECOND
FIGURE 3.30
TYPICAL RATE OF REMOVAL OF CURRENT (di/dt) EFFECT UPON
COMMUTATING dv/dt
3.11 GATE CIRCUIT RATINGS
Maximum ratings for the gate circuit are discussed in Chapter 4.
3.12 HOLDING AND LATCHING CURRENT
Somewhat analogous to the solenoid of an electromechanical
relay, an SCR requires a certain minimum anode current to maintain
it in the "closed" or conducting state. If the anode current drops below
this minimum level, designated as the holding current, the SCR reverts
to the forward blocking or "open" state. The holding current for a
typical SCR has a negative temperature coefficient; that is, as its junction temperature drops its holding current requirement increases.
This increase in both holding and latching current may be limiting in
military applications where -65°C operation is required. THE
DESIGNER IS URGED TO TAKE SPECIAL PRECAUTIONS TO
INSURE AGAINST LATCHING AND HOLDING CURRENT
PROBLEMS AND BY CONSULTING THE FACTORY WHERE
DOUBT MAY EXIST.
A somewhat higher value of anode current than the holding current is required for the SCR to initially "pickup." If this higher value
of anode latching current is not reached, the SCR will revert to the
blocking state as soon as the gate signal is removed. After this initial
pickup action, however, the anode current may be reduced to the
holding current level. Where circuit inductance limits the rate of rise
67
SCR MANUAL
of anode current and thereby prevents the SCR from switching solidly
into the conducting state, it may be necessary to make alterations in
the circuit. This is discussed further in Chapter 4.
A meaningful test for the combined effects of holding and latching
current is shown in Figure 3.31. The SCR under test is triggered by a
specified gate signal, under specified conditions of voltage, anode current, pulse width .and junction temperature.
The test circuit allows the SCR to latch into conduction at a current level I F1 • The test circuit then reduces the current to a continuously variable level I F2 • The current IF2 at which the SCR reverts to
the off-state is the desired value of holding current. See Chapter 20
for details of a suitable test circuit.
-lFI
\________
IF2}
''-______
FIGURE 3.31
TEST CIRCUIT
SETS VARIABLE
LEVELS OF IH
HOLDING CURRENT TEST WAVEFORM
3.13 REVERSE RECOVERY CHARACTERISTICS
During commutation SCR's display a transient reverse current that
far exceeds the maximum rated blocking current. This reverse current
is called reverse recovery current and its time integral is termed recovered charge. Figure 3.32 defines the salient reverse recovery parameters.
The cross-hatched area represents a common industry method of defining recovered charge (QRR), along with a method for defining recovery
time (trr). T4 is arbitrarily chosen to occur at intersection of the dotted
line drawn from iR through the iR/4 point, intersecting with the zero
current value. Thus defining recovery time as T 4 - T 1. Attempts at
using lower values than (iR/ 4) for the definition run into the problem
of measuring T 3 accurately for very soft recovery devices where the
recovery current slope may be very gradual.
Recovered charge is often specified in preference to trr due to its
strong application orientation. Specifically where the voltage across the
device must be limited by an R-C snubber network in series applications, the size of the capacitor required is determined by the SCR's
recovered charge characteristics (see Chapter 6 for details).
iT
0::
<.>
'"
FIGURE 3.32
68
SCR RECOVERY WAVEFORM DEFINITION
RATINGS AND CHARACTERISTICS OF THYRISTORS
Figure 3.33 shows an SCR's typical recovered charge characteristics.
100
8
60
40
"...... I-
20
~
§
10
...-:::: "...... VV ~
b3 ~~
, 8~
a:
'l
"......
:::::
---=
::::::::: -~ ;:.---
"......
I-~
~/ "......1-,...............
./'"
"......
--.----
.......
" 6 ..... , / ' ",..., .........V
V
...........
~
500
~ f-- 400
~~
---- ---
--
L-- 300
~ f--
200
100
~ I-- i
~~~o
-
'5
u
•
REVERSE
eII/d!
-
2
I
2
FIGURE 3.33
4
6
8 10
20
REVERSE di/dl - Alp. SEC
40
60
80 100
TYPICAL RECOVERED CHARGE (125°C) C158 SCR
It is to be noted that both QRR and 4. are strongly circuit dependent as well as device dependent. Both the peak-on-state current prior
to commutation as well as the commutation dil dt are significant circuit
variables. Additionally recovered charge has a positive temperature
coefficient requiring a fixed junction temperature as part of the test
conditions.
REFERENCES
1. ''Power Semiconductors Under Transient and Intermittent Loads,"
F. W. Gutzwiller and T. P. Sylvan, AIEEE Transactions, Part I,
Communications and Electronics, 1960, pages 699-706. (Reprint
available as Application Note 200.9. *)
2. "Take the Guesswork Out of Fuse Selection," F. B. Golden, Electronic Engineer, July 1969. (Reprint available as publication
660.21.*)
3. "Thermal Mounting Considerations for Plastic Power Semiconductor Packages," R. E. Locher, General Electric Application
Note 200.55. *
4. "Better Utilization of SCR Capability With AC Inductive Loads,"
J. C. Hey, EDN, May 1966, pp. 90-100. (Reprint available as Publication 660.12. *)
69
SCR MANUAL
5. "The Computerized Use of Transient Thermal Resistance to Avoid
Forward Biased Second Breakdown in Transistors," R. E. Locher,
Proceedings of the National Electronics Conference, Vol. 26, pp.
160-171, December 1970. (Reprint available as Publication
660.22.*)
6. "Ratings and Applications of Power Thyristors for Resistance
Welding," F. B. Golden, IEEE Industry & General Applications
Conference Record, #69C5-IGA, pp. 507-516.
7. "The Ratings of SCR's When Switching Into High Currents,"
N. Mapham, IEEE CP63-498, Winter General Meeting, New York,
N. Y., January 29, 1963. (Reprint available as Application Note
200.28.*)
8. "Behavior of Thyristors Under Transient Conditions," 1. Somos
and D. Piccone, Proceedings of the IEEE, Vol. 55, No.8, Special
Issue of High-Power Semiconductor Devices, August 1967, pp.
1306-131l.
9. "The Rating and Application of SCR's Designed for Switching at
High Frequencies," R. F. Dyer, IEEE Transactions of Industry
and General Applications, January/February 1966, Vol. ICA-2,
No.1, pp. 5-15. (Reprint available as Publication 660.13.*)
10. "The Characterization of High Frequency, High Current, Reverse
Blocking Triode Thyristors for Trapezoidal Current Waveforms,"
R. E. Locher, IEEE Transactions of Industry and General Applications, April 1968.
11. "The Rating and Application of a Silicon Power Rectifier," D. K.
Bisson, Rectifiers, in Industry, June 1957, publication T-93, Amencan Institute of Electrical ~ngineers, New York, N. Y.
12. "Bidirectional Triode Thyristor Applied Voltage Rate Effect Following Conduction," J. F. Essom, Proceedings of the IEEE, Vol.
55, No.8, Special Issue of High-Power Semiconductor Devices,
August 1967, pp. 1312-1317.
13. "Power Thyristor Rating Practices," J. S. Read, R. F. Dyer, Proceedings of the IEEE, Vol. 55, No.8, Special Issue on High-Power
Semiconductor Devices August 1967, pp. 1288-1300.
14. "Semiconductor Controlled Rectifiers-Principles and Applications
of p-n-p-n Devices," F. E. Gentry, et aI., Chapter 4, Prentice Hall,
Englewood Cliffs, N. J.
-Refer to Chapter 23 for availability and ordering information.
70
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
4
GATE TRIGGER CHARACTERISTICS, RATINGS,
AND METHODS
The ability of the triode thyristor (SCR or triac) to switch from
nonconducting to conducting state in response to a small control signal
is the key factor in its widespread utility for control of power. Proper
triggering of the thyristor requires that the source of the trigger signal
should supply adequate gate current and voltage, without exceeding
the thyristor gate ratings, in accordance with the characteristics of the
thyristor and the nature of its load and supply. The trigger source impedance, time of occurrence and duration of the trigger signal, and
off-state conditions are also important design factors. Since all applications of thyristors require some form of triggering, this chapter is
devoted to the fundamentals of the gate triggering process, gate characteristics and ratings, interaction with the load circuit, characteristics
of active trigger-circuit components, and basic examples of trigger circuits. This chapter will be devoted mostly to SCR's, while Chapter 7
contains more details on triac triggering. Specific trigger circuits for
performing various control functions are shown in subsequent chapters.
4.1 THE TRIGGERING PROCESS
Section 1.3 of Chapter 1 and Section 7.1.3 of Chapter 7 describe
the two-transistor analogy of the SCR, the junction gate and remote
gate operation of the triac, and the remote-base transistor action of the
SCR. From those discussions, it can be seen that the transition of a
thyristor from the non-conducting to the conducting state is determined
by internal transistor-like action.
The switching action, with slowly increasing DC gate current, is
preceded by symmetrical transistor action in which anode current
increases proportionally to gate current. As shown in Figure 4.1, with
a positive anode voltage, the anode current is relatively independent
of anode voltage up to a point where a form of avalanche multiplication
causes the current to increase. At this point, the small-signal (or instantaneous) impedance (dV /dI) of the thyristor changes rapidly, but
smoothly, from a high positive resistance to zero resistance, and thence
to increasing values of negative resistance as increasing current is
accompanied by decreasing voltage. The negative resistance region
continues until saturation of the "transistors" is approached, wherein
the impedance smoothly reverts from negative, to zero, to positive
resistance.
The criteria for triggering depends upon the nature of the external
anode circuit impedance and the supply voltage, as well as the gate
current. This can be seen by constructing a load line on the curves of
Figure 4.1, connecting between the open-circuit supply voltage, VL ,
71
SCR MANUAL
and the short-circuit load current, I A • With zero-gate current, the
thyristor characteristic curve intersects the load line at a stable point
(1). At a gate current of 1Gb the characteristic curve becomes tangential
to the load line at a point (2) where the negative resistance of the
thyristor is equal in magnitude to the external load resistance. Since
this condition is unstable, the thyristor switches to the low-impedance
state at stable operating (3). The gate current may now be removed and
conduction will be maintained at point (3). If the supply voltage is
reduced to VL2 the load line will shift and the operating point (3) will
'move toward the origin. When the load line becomes tangential to the
characteristic curve at point (4), the condition is again unstable, and
the thyristor reverts back to the high-impedance "off state."
The anode current at point (4) is the "holding" current for this set
of conditions. If, instead of reducing supply voltage to reach point (4),
the load resistance were increased, the point (5) at which the characteristic curve becomes tangential to the load line occurs at a lower
current, which is the holding current for that set of conditions. If the
gate current IGl were maintained while supply voltage was reduced
to VL3 , turn-off would have occurred at point (6), at a lower anode
current. A higher gate current, IG2 would then be required to trigger
the SCR, but reduction of this gate signal below IGl would allow it to
switch off, hence the SCR would not have been truly latched in the
on-state. The latching current is at least as high as the holding current
(at IG= 0), and is higher in some SCR's because of non-uniform areas
of conduction at low currents. In those cases, the triggering criterion
is not only meeting a negative-resistance intercept condition such as
point (2), but also reaching a certain minimum anode current at
point (3).
+ I
QUADRANT I
ANODE
CURRENT
-v
A NODE VOLTAGE
QUADRANT ]]I
-1
FIGURE 4.1
SCR ANODE-CATHODE CHARACTERISTICS WITH GATE CURRENT
The triac gate characteristics in quadrants I and III appear similar
to that of the SCR in quadrant I. It should be remembered that the
triac can be triggered with either a positive or negative gate signal
72
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
but that the tum-on process will not be perfectly symmetrical for all
possible biasing and triggering conditions.
Thyristor triggering requirements are dependent on both anode
and gate conditions. Therefore, specifications on a given thyristor's
requirements for gate voltage and gate current to trigger (VGT and I GT)
also define the anode circuit voltage and load resistance conditions.
4.2 SCR GATE·CATHODE CHARACTERISTICS
Trigger circuits must be designed to produce proper current How
between the gate and cathode terminals of the SCR. The nature of the
impedance which these two terminals present to the trigger circuit is
a determining factor in circuit design.
From basic construction and theory of operation, it can be seen
that the electrical characteristics presented between the gate and
cathode terminals are basically those of a p-n junction-a diode. This
is not the whole story.
4.2.1 Characteristics Prior to Triggering
Figure 4.2 shows the low-frequency full and simplified equivalent
circuits of the gate-to-cathode junction with no anode current Howing
(open anode circuit) for both conventional as well as for amplifying
gate SCR's. The series resistance RL represents the lateral resistance
of the p-type layer to which the gate terminal is connected. The shunt
resistance Rs represents any intentional or inadvertent "emitter short"
that may exist in the structure. The magnitudes of RL and Rs are variables resulting both from structure design and manufacturing process.
For example, Rs is extremely high in the C5 type SCR and quite low
in the C180 type which features emitter "shorts" to increase its VDRM
rating and dv / dt characteristic. The diodes are shown as avalanche
("zener") diodes because the reverse avalanche voltages of SCR gate
junctions are typically in the range from 5 to 20 volts, a condition easily
encountered in trigger circuits.
(0 1 FULL CIRCUIT
(b 1 SIMPLIFIED CIRCUIT
FIGURE 4.2(a)
GATE·CATHODE EQUIVALENT CIRCUIT FOR THE CONVENTIONALSCR
73
SCR MANUAL
'
GATE
~
ANODE
CATHODE
(1) Device Equivalent Circuit
(2) Gate Equivalent Circuit
FIGURE 4.2lb) GATE·CATHOOE EQUIVALENT CIRCUIT FOR THE AMPLIFYING GATE SCR
REVERSE
AVALANCHE
VOLTAGE
!
FIGURE 4.3 GATE'CATHODE CHARACTERISTIC CURVE II.. = 0)
The difference between a typical gate characteristic and an ordinary diode junction is shown in Figure 4.3. The relative effects of RL
and Rs are apparent in different regions of the curve.
The equivalent circuit and characteristics shown here are valid
only when anode current is zero or small as compared with gate current.
This information is, therefore, useful for reverse gate bias, for very
low forward gate current, and for examination of trigger circuits with
anode disconnected.
4.2.2 Characteristics at Triggering Point
With the anode supply connected, the equivalent gate circuit must
be modified, Figure 4.4, to include the anode current How across the
gate junction. Since anode current is a function of gate current (see
Chapter 1), the total current through the junction and the voltage drop
across the junction will increase more rapidly than with gate drive
alone. As anode current increases (Figure 4.5), the small-signal impedance between the gate and cathode terminals changes smoothly
from positive, to zero, to negative resistance. When the characteristic
curve becomes tangential with the load line of the gate signal source
impedance at point (1), the anode current becomes regenerative and
the SCR can then trigger. For specification purposes, "IGT" is the
maximum gate supply current required to trigger, hence is measured
at the peak of the curve (refer to Chapter 20 for test method).
Thus it is apparent that the impedance of the gate signal source is
another factor in the criteria for thyristor triggering.
4.2.3 Characteristics After Triggering
After the thyristor has been' triggered and anode current How
74
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
GATE SOURCE IMPEDANCE LOAD LINE
GATE
t
VG
CATHODE
RS
.,..!----.. . . .-----'
--r=-----v+-GT- - - - ' - - " ' - - VG
GATE VOLTAGE
FIGURE 4.4 GATE·CATHODE
EQUIVALENT CIRCUIT [1.0 = f(IGll
FIGURE 4.5 GATE CHARACTERISTICS,
ANODE CONNECTED
across the gate-cathode junction is sufficient to maintain conduction,
the gate impedance changes. Figure 4.4, shows that it behaves like a
source, having a voltage equal to the gate-cathode junction drop (at the
existing anode current) and an internal impedance R L. This voltage is
very nearly equal to the voltage drop between anode and cathode. The
characteristics under this condition are shown in Figure 4.6. The curvature in the fourth quadrant is effectively the result of an increasing RL
as more current is taken out of the gate. This is the result of the distributed nature of the gate junction as shown in Figure 4.2. As the
gate-to-cathode terminal voltage is reduced by withdrawing current,
the current How through the lateral resistance of the p-type layer causes
current to cease Howing through that portion of the p-n junction nearest
the gate terminal. This causes an increase in current density in areas
remote from the gate terminal. The higher current density and power
dissipation in the lateral resistance can cause thermal damage to the
thyristor.
+
---- ---IG
H
FIGURE 4.6
GATE CHARACTERISTICS AFTER TRIGGERING
If two SCR's are connected with gates and cathodes common, the
gate voltage produced by conduction of one SCR can, in some cases,
produce adequate triggering current in the gate of the other SCR.
In many instances, this may be a desired effect-turning both SCR's on
75
SCR MANUAL
simultaneously. In other cases, however, as when the anode supply
voltages of the two are 180 degrees out of phase, the existence of gate
current in the reverse-biased SCR can cause triggering at the instant
it becomes forward-biased because of stored charge in the p-type layer.
It can also cause excessive reverse current by the remote-base transistor
action.
4.3 EFFECTS OF GATE-CATHODE .IMPEDANCE AND BIAS
The preceding sections have shown that the criteria for triggering
involves the gate current, gate signal source impedance, and anode
supply (load) impedance. The interaction between gate and anode circuits demands examination in some depth.
4.3.1 Gate-Cathode Resistance
The two-transistor analogy shows that a low external resistance
between gate and cathode bypasses some current around the gate junction, thus requiring a higher anode current to initiate and maintain
conduction. Low-current, high sensitivity SCR's are triggered by such
a low current through the gate junction that a specified external gatecathode resistance is required in order to prevent triggering by thermally generated leakage current. This resistance also bypasses some
of the internal anode current caused by rapid rate-of-change of anode
voltage (dv/dt, see Chapter 3). It raises the forward breakover voltage
by reducing the efficiency of the n-p-n "transistor" region, thus requiring a somewhat higher avalanche multiplication effect to initiate triggering. The latching and holding anode currents are also affected by
the current which bypasses the gate junction.
The relative effect of the external resistance is dependent upon
the magnitudes of the internal resistances, RL and Rs of Figure 4.2.
For low-current thyristors, the type of construction used generally leads
to high values of Rs (virtually no emitter shorting) and low values of RL
because of the small pellet size. Figure 4.7 shows the effect of external
gate-to-cathode resistance upon holding current for the type CI06
low-current SCR. The spread between maximum and minimum values
represents production variations of the internal resistances and variations in current-gain of the equivalent "transistor" regions.
External shunt gate resistance also slightly reduces the turn-off
time of the SCR by assisting in recovering stored charge, by raising the
anode holding current, and by requiring higher anode current to initiate
re-triggering.
76
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
100
80
60
40
20
10
8
[3
...
II:
6
Il.
::I!
..
:3
~
~ ~~
"'""
NOTES: (I) CURVES SHOWN ARE FOR VARIOUS
JUNCTION TEMPERATURES.
K
""
........
V
~;"-..
.......
I"~ I'-..
MAXIMUM AT
::>
u
~
I
2
II:
I
..J
~ 0.8
0.6
0.4
0.2
--
....
FIGURE 4.7
I
-40°C
-r
I
I~ MAXIMUM AT 25°C - I--
~
~
IIO~ ~
~ t--.....
~
I
200
f....
t--.....f.....
f':
""'" t--..
MINIMUM AT -40°C
r-- K~~~~
r--- r-.
r-....
~
I
~
MINIMUM AT 110° C-
100
tT
r---- I--I'---r-...
I'---
r---.... r--
O. I
MAXIMUM
~
t--
I--
VOLTAGE RATING DOES NOT APPLY FOR
GATE TO CATHODE RESISTANCES GREATER
THAN 1000 OHMS.
~
...:z:I
Q
(3) CAUTION: STANDARD FORWARD BLOCKING
~~
4
~
~
(2) ANODE SUPPLY VOLTAGE = 12 VOLTS.
t\
!
400 600 800 1000
2000
4000 6000
10000
GATE TO CATHODE RESISTANCE-OHMS
SOOO
MAXIMUM AND MINIMUM HOLDING CURRENT VARIATION WITH EXTERNAL
GATE·TO·CATHODE RESISTANCE FOR C106 SCR
4.3.2 Gate·Cathode Capacitance
A low shunt capacitive reactance at high frequencies can reduce
the sensitivity of a thyristor to dv/dt effects (see Chapter 3), in much
the same manner as a resistor, while maintaining higher sensitivity to
DC and low frequency gate signals. This integrating effect is particularly useful where high-frequency "noise" is present in either the anode
or gate circuits.
77
SCR MANUAL
At the point of triggering, however, the gate voltage (see Figure
4.5) must increase as anode current increases. Therefore, a capacitor
connected between gate and cathode will tend to retard the triggering
process, yielding longer delay-time and rise-time of anode current. This
action can be detrimental when a high dil dt of anode current is
required (see Chapters 3 and 5 ).
After the SCR has been turned on, the gate acts as a voltage
source, charging the capacitor to the voltage drop across the gate junction. Since this voltage (depending on value of anode current) is generally higher than the gate voltage required to trigger the SCR (VGT),
the energy stored in the capacitor can supply triggering current for a
period of time after removal of anode current, thereby possibly causing
the SCR to fail to commutate. In low-current SCR's, a capacitor on the
order of 10 microfarads can maintain gate current for over 10 milliseconds, hence can prevent commutation in half-wave, 50 or 60 Hz
circuits.
If the gate triggering signal is a low-impedance pulse generator
in series with a capacitor, the capacitor can be charged by gate current
during the pulse and the pOlarity will be such that at the end of the
pulse the SCR gate will be driven negative. For low values of anode
current at this instant, the negative drive may raise the holding current
requirement above the anode current and tum off the SCR.
4.3.3 Gate·Cathode Inductance
Inductive reactance between gate and cathode reduces sensitivity
to slowly changing anode current or gate source current while maintaining sensitivity to rapid changes. This differentiating effect is useful
in improving thermal stability since changes in thermal leakage current
are slow. When used with the light-activated SCR, it provides sensitivity to a Hash of light with insensitivity to steady-state ambient light
(see Chapter 14).
With anode current Howing, the gate voltage causes current to
How out of the gate, through the inductance. The rate at which this
current builds up after triggering is a function of the L/R ratio of the
inductance to both internal and external resistance. As this negative
gate current rises, the holding current of the thyristor also rises. If
anode current is low, or increasing more slowly than negative gate current, the thyristor may drop out of conduction.
After the SCR anode current ceases, negative gate current will
continue for a period of time, decaying according to the L/R timeconstant. This negative gate current during the tum-off condition can
reduce tum-off time (by nearly 10:1 in small SCR's) and can permit a
faster rate of re-applied off-state voltage (higher dv/dt).
If a triggering current pulse is applied in parallel with an inductor
and the gate, the pulse can produce a current How through the inductor.
At the termination of the pulse, the inductor current will continue to
How as a negative gate current, thereby raising holding current and
possibly causing tum-off of the SCR.
78
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
4.3.4 Gate-Cathode LC Resonant Circuit
A parallel LC resonant circuit connected between gate and cathode
can provide a frequency-selective response, and can also produce a
condition of oscillation.
The oscillating condition is obtained by making the anode current
0) holding current and
value intermediate between the normal (IG
the holding current with maximum negative gate current flowing
through the inductor. As explained in Section 4.3.3, the SCR can be
turned on, then negative gate current will increase until the SCR turns
off. After turn-off, inductor current will charge the capacitor to a negative voltage, then the capacitor will discharge into the inductor in a
resonant manner. When the capacitor voltage swings positive again, it
can re-trigger the SCR and the process will repeat indefinitely. Damping is required to avoid such oscillation.
=
4.3.5 Positive Gate Bias
The presence of positive current in the gate when reverse voltage
is applied to the anode may increase reverse blocking (leakage) current
through the device substantially. As a result, the SCR must dissipate
additional power. Therefore it is necessary either to make provision for
this additional loss or to take steps to limit it to a negligible value.
Figure 4.8 gives the temperature derating for different SCR lines
at various gate drive duty cycle (percent of full cycle or 360 electrical
degrees) for values of peak positive gate voltage. For proper application, this loss must be included in the total device dissipation. The
temperature derating, AT, found from Figure 4.8, must be subtracted
from the maximum allowable stud temperature (found from the device
rating curve) for the proper cell type and conduction angle. For lead
mounted devices, subtract from the ambient temperature curve. Derating becomes negligible if the gate voltage is less than 0.25 volt or the
temperature derating turns out to be lOC or less.
79
SCR MANUAL
;;.,.;
100
90
80
70
....v
60
1/
50
C 35 AND C36 SERIES
/1
40
/
/
30
(
I
V
V
./
V
./
CIOAND CII_
SERIES
I
I
V
I
/
Ir'~
7
II
7
I
C50
SERIES
~
I
/
I 7
/
17
71 7
7
~
..............
V
//" I
I
j
3
~J
2
I
V
C8SE~
o
FIGURE 4.8
7
CURVES
' " 'OTHER
FOR GATE
CYCLE OF"'OWN
50°/0. FOR
GATEDUTY
DUTY
CYCLE MULTIPLY b. T VALUE BY
FOLLOWING FACTORS:
DUTY CYCLE, % FACTOR DUTY CYCLE, % FACTOR
33
0.67
16.5
0.33
25
0.50
8.3
0.17
0.5
1.0
1.5
2.0
2.5
3.0
3.5
PEAK POSITIVE GATE VOLTAGE-VOLTS
TEMPERATURE DERATING CURVE FOR SIMULTANEOUS APPLICATION
OF POSITIVE GATE PULSE WHEN ANODE IS NEGATIVE
A means of limiting the additional reverse dissipation to a negligible value is given by a gate clamping circuit of the type shown in
Figure 4.9 for low and medium current SCR's (ClO and C35 series).
Resistor RA and a diode are connected from gate to anode to attenuate
positive gate signals whenever the anode is negative. For a given peak
value of open circuit gate source voltage, Figure 4.9 gives the maximum ratio of the value of RA to RG that will safely clamp the gate for
all values of reverse voltage within the reverse voltage rating of the
SCR.
An alternate way to limit additional reverse leakage dissipation
due to positive gate voltage is to insert in series with the SCR a recti-
80
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
1000
\\
Rl
\
GATE SIGNAL A~
SOURCE
<
I
I
RS I~;
I
I E
I
\
r~,
\
~:'
L ______ .J
100
G~ICIRCUIT
~
(
,ow
IMPEDANCE
TO GATE
SIGNAL REaD)
..
~:,. TERMINALS
PEAK V(OPEN CKT. TERMINAL
VOLTAGE OFSOURCEI
G=PEAK I (OUTPUT CURRENT OF SOURCE
WITH SHORT CKT ACROSS TERMINALS
"10
2
FIGURE 4.9
*
TO
POWER
<'>
4
~ ........
"-. r-......
6
8
PEAK E (VOLTS)
............
10
12
GATE CLAMP CIRCUIT FOR CONTROLLED RECTIFIER
fier diode that has a lower reverse blocking current. In this manner the
diode will assume the greater share of the reverse voltage applied to
the series string, significantly reducing reverse dissipation in the SCR.
4.3.6 Negative Gate Bias
The gate should never be allowed to become more negative with
respect to the cathode than is indicated on the specification bulletin.
For example, the gate of the C35 (2N681) type has a rated peak reverse
voltage of 5 volts. If there is a possibility that the gate will swing more
negative than the rated value, a diode should be connected either in
series with the gate, or from cathode to gate to limit the reverse gate
voltage. A considerable negative gate current (conventional current
flow out of the gate) can be caused to flow if the cathode circuit between cathode and gate is opened for any reason while the SCR is
conducting forward load current (conventional current flow from anode
to cathode). This current would initially be limited only by the impedance of the gate circuit and could cause the allowable gate dissipation
to be exceeded, thus leading to possible failure of the SCR.
When the anode is positive, negative gate bias tends to increase
the forward breakover voltage V(BO) (Section 1.9.1) and the dv/dt
withstand capability (Section 1.5) at a given junction temperature for
small SCR's without internal emitter shorting. For example, the C5
types (2N1595, C106, etc.) have VDRM specified for a certain value of
81
SCR MANUAL
gate-to-cathode resistance (RGK = 1000 ohms) and at a specified junction temperature. For more detail on the effect of negative gate bias
on small SCR's the reader is referred to Reference 1.
l
.1..
IFXM
.J....
TRIGGE R ct---_-~Io(,J,./
TRI GGER O---'-->o(,..j......o'
SIGNAL
CRI
SIGNAL
GEIN5059
+
(0 I VOLTAGE
BIAS
+
(bl CURRENT BIAS
FIGURE 4.10 NEGATIVE GATE BIAS ARRANGEMENTS
Figure 4.l0(a) shows a voltage bias arrangement. Resistor Rb is
taken to a negative supply instead of being merely returned to the
Eb- D
cathode. The voltage source Eb establishes a current Ib """
R ' where
D is the voltage drop across diode CRl (typical value 0.7 volt). The
diode provides a fixed negative bias voltage gate-to-cathode for the
SCR. The disadvantage of this approach, however, is the loss of input
sensitivity due to resistor R b.
Figure 4.l0(b} shows a current bias scheme useful for smaller
junction diameter SCR's. Resistor Rb and the bias source are selected
so that a bias current Ib """ I FxM is established through resistor Rb in
the direction indicated; I FxM is the maximum forward blocking (leakage) current of the SCR under the prevailing junction temperature
and anode voltage. Selection of Ib in this manner yields a "worst case"
design on the assumption that most, if not all, of IDRJ\r will be diverted
from the SCR emitter (gate-cathode junction). This approach is limited
to SCR's which have sufficient reverse gate power ratings to handle
reverse current Ib at its associated reverse gate voltage. The scheme
of Figure 4.l0(b} is suitable, for example, for General Electric C5 type
SCR's which allow operation of the gate-to-cathode junction in reverse
avalanche.
The improvement in dv / dt withstand capability that can be
achieved by negative gate biasing is shown in Figure 4.11 for a typical
C35 type SCR. It shows the effect of gate bias on the allowable time
constant of application of forward blocking voltage without having
the SCR switch on. The zero gate voltage curve corresponds to the time
constant values given on the C35 specification sheet for the open gate
condition. Figure 4.11 extends the usefulness of this information for
different values of gate bias.
82
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
C!-
C35E
GATl~ onI
VO T GE
0
q'
,.;
I
N
I
I
0
0
I
I
/
v
e35H
C35B
C35G
C35A
C35F
C35U
o
2
4
6
0
17
v
/
I
/I
I II / / V /
/ I /, '1 I /
/ I/, / /
/ II/; '/ V V V
/~ If/ / / . / v
///h 1/.V /
If/I/, / . /
C35C
.l,
(\1-
0
0
II I II II V /
j /
II V J
C35D
~~- ~l
~
+
/'"
V
/
.//
./
8
./
V
10
12
14
16
18
TIME CONSTANTS (t'l-MICRO SECONDS
FIGURE 4.11
EFFECT OF GATE BIAS ON ALLOWABLE TIME CONSTANT OF
APPLICATION OF FORWARD BLOCKING VOLTAGE
It is possible to design circuits which apply a negative gate bias
or short the gate to the cathode only while dv/dt is being applied.
These circuits do not degrade the gate signal as much as in Figure 4.10
but are expensive for most applications.
The basic idea is to differentiate the dv I dt applied to the anode,
invert the polarity and apply it to the gate. Figure 4.12 shows a transistorized dynamic snubber. RIC supplies base current to QI turning it
on when anode voltage is rising. Gate triggering would be lost during
the time of rising dv/dt since the gate is being shunted. However, the
insertion of Q2 avoids this problem since now the gate signal not only
triggers the SCR but first shunts the base drive to QI' Both QI and Q2
should be epitaxial transistors with low saturation voltages.
r---J4---,I
I
I
I
I
I
I
C
GATE~~------+---~~------~--~
FIGURE 4.12
TRANSISTOR SNUBBER TO IMPROVE dv/dt
83
SCR MANUAL
Large area devices and devices with emitter shorting are influenced little by gate shorting because of the· shunting effects of Rs
(Figure 4.2(a)). Unless V (RO) is specified with a bias resistor, conservative circuit design practice should not depend upqn increasing V (BO)
by negative gate biasing. Some types of SCR's that feature n-type gate
structures (C501, C601, etc.) as well as triacs can be triggered by either
positive or negative gate signals. Under no circumstances should negative gate vias be used with these types to enhance blocking stability.
The influence of turn-off time by different gate bias techniques
seems to be very limited because it is mainly a function of carrier lifetime in an area not accessible by the gate.
4.4 EFFECTS OF ANODE CIRCUIT UPON GATE CIRCUIT
In Section 4.1 it was shown that the anode circuit voltage and
impedance were determining factors in triggering. The effect of anode
current was discussed in Section 4.2.3. Two other effects are worth
noting. Junction capacitance in the SCR can couple high-frequency
signals from the anode to the gate circuit which, although they may
not cause triggering in themselves, may interfere with normal operation
of the trigger circuit.
When the anode voltage of the SCR reaches either the forward
breakover or reverse avalanche voltage, a voltage will appear at the
gate terminal. In the case of forward breakover voltage, a forward
anode current starts flowing which produces a positive gate voltage,
as in normal conduction (see Section 4.2.3). When the reverse avalanche
voltage is reached, the gate junction becomes reverse biased. Depending on the magnitude of Rs (Figure 4.2) the negative voltage appearing at the gate terminal may rise to the avalanche voltage of the gate
junction. If a reverse voltage transient on the anode exceeds reverse
avalanche, the reverse-blocking junction of the SCR no longer blocks,
thereby applying the transient energy to the gate junction in reverse.
The gate junction and any external circuit connected to the gate may
then receive excessive voltage and current from this process.
When the SCR is conducting, its gate is essentially at the same
potential as its anode. When the SCR is non-conducting, the gate
potential is not related to anode potential within the normal operating
range. However, during the commutating transition from conduction
to non-conduction, the gate goes through an intermediate phase which
can result in a large negative voltage appearing at the gate terminal.
If an SCR is commutated, as in a DC chopper or Hip-flop circuit, by
the step application of a reverse bias, the gate voltage will initially be
the normal forward gate-cathode junction drop until that junction
recovers, whereupon both anode and the gate will go negative. The
gate voltage will then follow anode voltage until the main reverseblocking (p-n) junction recovers, at which time the gate reverts to its
normal characteristics. These transitions are readily observed on small
SCR's in particular. On larger SCR's, the effects are somewhat masked
by lower values of internal shunt resistance Rs. The negative transient
at the gate can cause malfunction or damage in the external gate circuit.
84
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
4.5 DC GATE TRIGGERING SPECIFICATIONS
I.
TRIGGER ALL UNITS AT:
~
I
~3
* /:?
14
\1 2
I~
10
+20·C -6'·C
~
~
;.J
•
~
..
- 60·C
MIN GATE VOLTAGE
REQUIRED TO
TRIGGER ALL
~,~
~
~ 0 ~ ~ GERANYUNITSAT
~1~lV:rr~T::J~
~
12S-C ~ 0.25 V
;
°0
(AI
(01\
~:l:~=~US
a
50
100
5.0
,NO~S::I) C~SE T~MP~AT~RE .'
,
121 S~J<'A~J~r~'EPRE-
~~
WATTS
SENT LOCUS OF POSStBLE
,
o
o
PREFERRED GATE
DRIVE AREA
M
M
The DC gate trigger characteristics of an SCR are presented in
the fonn of a graph similar to Figure 4.13 which applies to the C35
(2N681) type SCR. The graph shows
gate-to-cathode voltage as a function of positive gate current (How
from gate to cathode) between limit
lines (A) and (B) for all SCR's of
the type indicated. These data apply to a zero-anode-current condition (anode open).
!~;.~E~~+~5~~~ FROM
"-
....
~
~
-
INSTANTANEOUS GATE CURRENT-mA
MAX ALLOWA.l.....
DISSIPATION
uNiTS
I I
I I
~ /0.~ ~
"""\ !z
±
I I
MIN GATE CURRENT REQUIRED
(.,
~
...._.,
I I
I I
~
~
~
FIGURE 4.13 DC GATE TRIGGERING
CHARACTERISTICS (FOR C35 TYPE SCR)
INSTANTANEOUS GATE CURRENT - AMPERES
The basic function of the trigger circuit is to simultaneously supply the gate current to trigger IGT and its associated gate voltage to
trigger VGT' The shaded area shown in Figure 4.13 contains all the
possible trigger points (IGT' VGT) of all SCR's conforming to this specification. The trigger circuit must, therefore, provide a signal (IG, VG)
outside of the shaded area in order to reliably trigger all SCR's of that
specification.
This area of SCR gate operation is indicated as the "preferred
gate drive area." It is bounded by the shaded area in Fig'ure 4.13
which represents the locus of all specified triggering points (IGT' VGT),
the limit lines (A) and (B), line (C) representing rated peak allowable
forward gate voltage VGF, and line (D) representing rated peak power
dissipation Pmr . Some SCR's may also have a rated peak gate current
IGFM which would appear as a vertical line joining curves (B) and (D).
The insert in the upper right hand portion of Figure 4.13 shows
the detail of the locus of all specified trigger points, and the temperature dependence of the minimum gate current to trigger IGTmln' The
lower the junction temperature, the more gate drive is required for
triggering. (Some specifications may also show the effect of forward
anode voltage on trigger sensitivity. Increased anode voltage, particularly with small SCR's, tends to reduce the gate drive requirement.)
Also shown is the small positive value of gate voltage below which no
SCR of the particular type will trigger.
The reverse quadrant of the gate characteristic is usually specified
in terms of maximum voltage and power ratings. The application of
reverse bias voltage and the extraction of reverse gate current for
SCR off-state stability was discussed in Section 4.3.6.
4.6 LOAD LINES
The trigger circuit load line must intersect the individual SCR
85
SCR MANUAL
gate characteristic in the.region indicated as "preferred gate drive area"
in Figure 4.13;··The intersection, or maximum operating point, should
furthermore belocated'as,close to the maximum applicable (peak, average; etc.} gate power dissipation curve as possible. Gate current rise
times should be in the-order of several amperes per microsecond in the
interest-of minimizing anode tum-on time particularly when switching
into high currents. This in turn results in minimum turn-on anode
switching dissipation and minimum jitter.
Construction of a "load line" is
a convenient means of placing the
maximum operating point of the
trigger circuit-SCR gate combination into the preferred triggering
area. Figure 4.14(a) illustrates a
la) SATE CIRCUIT
basic trigger circuit of source voltage e. and internal resistance Rs
driving an SCR gate. Figure 4.14(b)
shows the placement of the maxiEQUIVALENT TRIGGER CIRCUIT
mum operating point well into the
"preferred trigger" area close to the
rated diSSipation curve. The load
'.Eo<:
t
line is constructed by connecting a
straight line between the trigger
'seR. cHARACTERISTIC
circuit open circuit voltage E"",
entered on the ordinate, and the
. trigger circuit· short circuit current
,.....
LOAD LINE
-- -
~
(
ISC"~ t;
LOCUS OF ALL SPECIFIED
TRIGGER POINTS
I bl L.OAD LINE
·s
SUPERPOSED ON GATE TRIGGER CHARACTElaSTIC
Isc
= ~:c
entered on the abscissa.
FIGURE 4.14 ·GATE CIRCUIT
AND CONSTRUCTION OF LOAD LINE
If the trigger circuit source voltage is a function of time e. (t), the
load line sweeps across the graph, starting as a point at the origin and
reaching its maximum position, the load line, at the peak trigger circuit
output voltage.
The applicable gate power curve is selected on the basis of
whether average or peak allowable gate power dissipation is limiting.
For example, if a DC trigger is used, the average maximum allowable
gate dissipation (0.5 watt for C35) must not be exceeded. If a trigger
pulse is used the peak gate power curve is applicable (for the C35, the
5 watt peak power curve labeled D in Figure 4.13). For intermediate
gate trigger waveforms the limiting allowable gate power dissipation
curve is determined by the duty cycle of the trigger signal according to:
peak gate drive power X pulse width X pulse repetition rate ~ allowable average gate power.
Inverter type SCR's that require a stiff gate signal because of high
di/dt and high frequency operation often have peak pulse gate power
curves (Figure 4.15). These pulse curves take advantage of the transient thermal resistance of the gate in order to achieve higher power
pulses. But again it must be remembered that the average gate power
dissipation should not be exceeded.
86
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
44
1--r---1c--;c--;---t~NO--'T-ES--':1-"1 1 I 1 1 1 1 1 I !
I-+--t---t----t-t- (I ) REQUIREMENTS SHOWN HEREON MUST BE MET WHEN ANODE CuRRENT RATE OF RISE
401-+---t-1A
rl----t-t11
1-+---1I/"f-/:'1:H----t-t-:
11..,...\
61-+---lJ-9t----t-t-
~
VI
32
We
~
....
28
6
~
~
10
PERMI SSIBLE-
GATE SOURCE LOAD LINES.
(3) MINIMUM PERMISSIBLE GATE DRIVE" 20 VOLTS
OPEN CIRCUIT, 20 OHMS. LOWER GATE DRIVE
MAY CAUSE DEVICE FAILURE.
-
Vt\.
V r\
1--+1ft--71-1-"1---tlf--+l./\
(~l
CASE TEMPERATURE
~
+25 D C T()
+ 120 DC
(6)
CuRVES SHOW MAX IMUM ALLOWABLE PEAK GATE POWER DISSIPATION VS. PULSE WIOTH_
(7)
MAXIMUM ALLOWABLE AVERAGE
DISSIPATION :I.DWArT
/'
24
~
~
~
~
1---I--IbAI---'f--+-
/~
>
l!'
~~:~:;SA::: :~;:~SCgg~~6 OF
(41 MINIMUM PERMISSIBLE GATE PULSE WIDTH:
~~'tSsEic~AXrMUM GATE PULSE RISE TIME,.
V
~
g
(2)
I
__
GATE POWER
_
iIi
I
t-- P" ~, >C-:Jr>\-\t----t---!----+-I-+---+---t---t--t- t----t-L
~~~~\+~-+--t-~__+-I-+-~I-t-_r_t__+~
v1-~~v"\
20
'6
1\
r-
V :.-t~ ~'.I'=
1-<·t~
/:\:::=:==:=:=-+--+--1--+-_
-' ;.
_1 ___
--+--+---+---1
j--
:~ ~,,~ ~V~';;"IS.tD->-L-+--~-+--+-+-f--t---1
"~
"'0 "{--
/rL,,--
100
~ ~o~:;..;><;:V /r/~1'>.. '-«8~-C-..,~+--+-
.2
DI
~ /~\ /t?5<~~VV·"3-p-"p.---f'-,--f!=.- ~::
-;.
L-
\«,~
PS:Vy V[./[>I>:::- 4.\: t-+--+--+--t---l
\
'i Ie:: l2 ~~
f::
t-'"'
I
I-- t--
oV
o
\\
0.8
i
1.6
2.4
I
3.2
4.0
INSTANTANEOUS GATE CURRENT -
FIGURE 4.15
4.8
5.6
6.4
AMPERES
GATE TRIGGER REQUIREMENTS FOR HIGH FREQUENCY
AND HIGH dl/dt OPERATION
4.7 POSITIVE GATE VOLTAGE THAT WILL NOT TRIGGER SCR
Figure 4.13 also indicates the maximum gate voltage that will not
trigger the SCR. For example, for the C35 (2N681) type, Figure 4.13
shows that at 125°C junction temperature this value is 0.25 volt. This
limit is important when designing a trigger circuit which has a standby
leakage current when no trigger signal is present. Examples of this are
saturable reactors and directly coupled unijunction transistor trigger
circuits. To prevent false triggering under these circumstances, a resistor should be connected across the output of the trigger circuit. Its
value of resistance in ohms should not exceed the maximum gate voltage that will not trigger divided by the maximum trigger circuit
standby current.
4.8 PULSE TRIGGERING
Thyristors are commonly specified in terms of the continuous DC
gate voltage and current required to trigger. For trigger pulse widths
down to 100 microseconds, the DC data apply. For shorter pulse
widths, VGT and IGT increase.
87
SCR MANUAl
On a short-time basis, thyristors may be generally considered to
be charge controlled, as are transistors. The free charge stored within
the gate p-type layer of an SCR may be considered to be the difference
between the incoming charge How rate (dq/ dt = I G ) and the internal
recombination rate. Under DC conditions and for a given recombination rate, the free charge is directly a function of gate current. When
the free charge reaches a certain level, the device triggers. To get the
required charge into the gate in a time that is short compared with
the recombination time requires higher current (hence higher voltage)
than for DC triggering.
100
0
•
.,..
6
4
10
•
Z
=::r_~
I
SLOPE
•
Q
O. 6
Q4
O. I
MAXIMUM GATE TRIGGER
CURRENT AT -40°C
I'
""
MAXIMUM GATE TRIGGER
CURRENT AT +Z5OC
I
I
2
I
0.8
k,
1,\
I ~AXIMUM
GATE TRIGGER
,..........
VOLTAGE AT -40~~
~~~~~~ ~~T~ ~~~GcGEV
0..
I
0.4
NOTE:APPLIES FOR RECTANGULAR TRIGGER PULSES.
0.2
0.1
III
LLJI'I
0.1
0.2
0.4 0.6
,
4
6 8 10
II
2':>
40 60 100 200 400 600 1000
GATE PULSE WIDTH - MICROSECONDS
( 0) CURRENT ANO VOLTAGE REQUIREMENTS
FIIIURE 4.1B EFFECT OF TRlIIIIER PULSE WIDTH (C·10B SCR)
Figure 4.16(a) shows the relationship between pulse width and
peak current for a rectangular pulse to trigger the C-106 type SCR.
Note that the current curves approach a constant-charge slope at the
smaller pulse-widths. The point at which the pulse current curve
departs from the DC current level is about 200 microseconds for this
small thyristor. Other SCR types, with shorter recombination times,
can be triggered with pulse current equal to the DC level down to
about 20 microseconds.
88
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
It should not be inferred from Figure 4.16(a) that only rectangular pulses are acceptable. Any unidirectional waveshape which does
not exceed gate current, voltage, and power ratings may be used if
the total charge is adequate. Proper charge criteria may be determined
by plotting, as in Figure 4.16(b), the integral of the actual current wave
and the integral of the rectangular pulse current. If the two curves
cross, the triggering charge is adequate.
Figure 4.17 shows the increase in gate drive required for triggering four types of SCR's with trigger signals of short pulse duration.
In order for the SCR to trigger, the anode current must be allowed to
build up rapidly enough so that the latching current of the SCR is
reached before the pulse is terminated. (Latching current may be
assumed to be three times the value of the holding current given on
the specification sheet.) For highly inductive anode circuits one must
use a maintained type of trigger signal which assures gate drive until
latching current has been attained.
1000
1'-.
"" ~
4
.5
fZ
...
"'-
0::
0::
::>
u
...
0::
C>
C>
"
100
~
~C-20
0::
...
f-
...........
!:i
C>
C-150
~ r--
- "'"
C-45
I-
I-NOTES:
\. CURVES SHOW MAXIMUM VALUES
FOR ALL SPECIFIED JUNCTION TEMPERATURES
2.RECTANGULAR PULSES WITH RISE TIMES LESS
THAN 100 n SEC.
I
10
10
100
GATE PULSE WIDTH (J.L SEC.l
FIGURE 4.17
GATE DRIVE REQUIRED FOR SHORT TRIGGER PULSE DURATION
One situation encountered frequently is that which exists when a
capacitor is discharged to provide a latching current pulse in a highly
inductive circuit. This situation is depicted in Figure 4.18 of on-state
current.
IZ
"'
"'
~
u
"'
~
I
Z
o
!-------TIME
FIGURE 4.18
CURRENT WAVEFORM FOR CAPACITIVE AND INDUCTIVE LOAD
89
SCR MANUAL
The latching pulse from the capacitor discharge is followed by a
slowly rising anode current determined primarily by inductance in the
circuit. Very often, confusion exists regarding holding current and
latching current in such cases. If the gate trigger pulse ends before
the end of the initial current pulse, the device must remain in the
on-state at the valley point where the main circuit takes over.
If the device has a holding current higher than the valley current
level provided, it will go out of conduction and the circuit will not
latch. This is, however, a result of high device holding current rather
than latching current. However, if the gate trigger signal lasts beyond
the valley point before it is ended and the device still fails to latch,
then it is a latching current problem rather than a holding current
problem.
The DC gate trigger characteristics are measured on a ~OO%
basis in production for all SCR's, but the pulse trigger characteristics
are measured only on a sampling basis. For applications where the
pulse trigger characteristics are critical, a special specification should
be requested so that satisfactory pulse triggering will be assured.
4.9 ANODE TURN-ON INTERVAL CHARACTERISTICS
Figure 4.19 shows the turn-on, or switching, characteristics of a
typical CI0 type SCR. It is representative of other SCR types as well.
Percent anode voltage is shown as a function of time, following application of the trigger signal at zero time, for switching from 500 volts
and from 100 volts for two different circuit current levels.
100~----~~~~~----------,------------,------------,
GATE SIGNAL APPLlEO AT ZERO
TI~r
FROM
~------++--t-lk-+----- FROM 7.0 VOLT (OPEN CIRCutTl, ;0 OHM
...'"
'"~
SOuRCE ,Ot MICROSECOND RISE TIME
...
'"x
60
0
...
...
...
'"
0
40
0
Z
o
"a.
I-
z
~ 20
~,
SCR "HALF ON"-
u
'"
'"
0:
'"a.
I
"-"-
I
10
.0
'"u
'"
0
o
20
40
60
R, POTENTIOMETER SETTING IN PERCENT
80
(b) Transfer Function
FIGURE 4.21
92
SIMPLE HALF·WAVE VARIABLE RESISTOR PHASE CONTROL
(LIMITED RANGE OF CONTROL)
10 0
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
extended beyond the 90 degree point because the trigger circuit supply voltage and the trigger voltage producing the gate current to fire,
Em, at the peak of the ACsupply voltI GF, are in phase. When e'l<'
age, the SCR can still be triggered with the maximum value of resistance between anode and gate. Since the SCR will trigger and latch
into conduction the first time IGT is reached, its conduction cannot be
delayed beyond 90 electrical degrees with this circuit.
The transfer function of this circuit has also been plotted in the
same figure. It assumes that the potentiometer R has been chosen so
that the SCR just does not fire at the maximum setting. The transfer
function is very non-linear and repeatibility of setting is not possible
either with different SCR's or with temperature due to IGT variation.
Figure 4.22 shows an R-C-Diode circuit giving full half-cycle control (180 electrical degrees). On the positive half-cycle of SCR anode
voltage the capacitor will charge to the trigger point of the SCR in a
time determined by the RC time constant and the rising anode voltage.
On the negative half-cycle, the top plate of the capacitor charges to
the peak of the negative voltage cycle through diode CR2, thus resetting it for the next charging cycle.
Since triggering current must be supplied by the line voltage
through the resistor, the capacitor must be selected such that its charging current is high compared with I GT , at the instant of the latest
desired firing angle. Conversely, select the maximum value of R to
produce IGT at the latest desired firing angle, using the line voltage
=
TYPICAL CIRCUIT VALUES FOR Eoc'" 120V
seR" GE CI06
R:: lOOK
C:: O.IJLF
CRI, CR2" IN5059
(a)
CONDITIONS:
IGT " 200,..
50
II:
OJ
-........
~
.. 40
....
..J
III
"
..J
~ 30
~~
I "~
:
~
"co
OJ
"
~ 20
OJ
.
I
u
a:
OJ
10
1.0
0.2
0.3
i
;
~i
I
0.4
0.5
0.6
POTENTIOMETER SETTING
i
i
0.7
~
0.8
0.9
I .0
(b) Transfer Function
FIGURE 4.22 SIMPLE HALF·WAVE RC-DIDDE PHASE CONTROL (FULL 180· CONTROL RANGE)
AND ITS ASSOCIATEO TRANSFER FUNCTION
93
SCR MANUAL
less IR drop in the load at that point, then select C to produce VGT at
that point in time. But similar to all simple RC triggering methods,
non-linear output results as the transfer characteristic (Figure 4.22)
shows. Again it should be reiterated that since the output depends so
heavily on I GT , it will vary with temperature and different devices.
Figure 4.23 illustrates a slave circuit arrangement in which an
independent half-wave circuit (SCR 2 ) is triggered on one half-cycle
at a predetermined phase angle. On the following half-cycle the slave
circuit will trigger SCR I at the same phase angle relative to that halfcycle. When SCR 2 does not trigger, capacitor C will charge and discharge to the same voltage at the same time constant. The voltage across
C will not be sufficient to trigger SCR I • As SCR 2 is triggered, capacitor
C on discharging sees a time integral of line voltage that is different
from the one on charging by the time integral of voltage appearing
across the load. This action resets the capacitor to a voltage level
related to the trigger delay angle of SCR 2 • On the next half-cycle, when
the anode of SCR I swings positive, it will trigger at the end of this
delay angle.
--.-------f-----1
1----,----,
LOAD
15K
01
-.GED2308
D2
15K
120 VAC
GE DT230B
03
C
SCRI
SCR2
5.0MFD
SLAVE
MASTER
seRI, SCR2: GE CII/C20 TYPES
FIGURE 4.23
THREE TERMINAL, FULL WAVE, RC·DIODE SLAVING CIRCUIT FOR
FULL·WAVE PHASE CONTROL
4.11 TRIGGERING SCR WITH A NEGATIVE PULSE
Some applications may make it desirable to trigger an SCR with
a negative pulse rather than with one of the conventional positive
polarity. In low power level SCR circuits a diode connected in series
with the SCR allows negative triggering conveniently and economically.
Figure 4.24 shows this arrangement for a C103 type SCR.
FIGURE 4.24
NEGATIVE PULSE TRIGGERING
A complementary SCR, like the C13, is designed for negative
voltage triggering. Therefore, this device should be used in low voltage
applications « 40 V) where negative voltage triggering is a must.
94
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
4.12 AC THYRATRON-TYPE PHASE SHIFT TRIGGER CIRCUITS
Figure 4.25 illustrates a full-wave phase controlled rectifier
employing an R-C or R-L phase shift network to delay the gate signal
with respect to the anode voltage on the SCR's. Many variations of
this type of phase shift circuit have been worked out for thyratrons.
AC
SUPPLY
_
DC
OUTPUT
FOR C
FIGURE 4.25
R·C OR R·L PHASE SHIFT NETWORK CONTROL OF SINGLE PHASE BRIDGE OUTPUT
When using SCR's (C8, CI0, Cll, C35, C36, and C50 series),
the following criteria should be observed to provide the maximum
range of phase shift and positive triggering over the particular SCR's
temperature range without exceeding the gate voltage and current
limitations:
A. The peak value of Ve should be greater than 25 volts.
_1_
2 fL< V e _ 9
27TfC or 7T = 2
where C = capacitance in farads
L = inductance in henries
Vc = peak end-to-end secondary voltage of control
transformer
f = frequency of power system
V - 20
C. Rs = "0.2
where Rs = series resistance in ohms
B.
D. Rc S ~~C or 10 (27TfL)
Because of the frequency dependence of this type of phase shift
circuit the selection of adequate L or C components becomes easier at
higher operating frequencies.
4.13 SATURABLE REACTOR TRIGGER CIRCUITS
Saturable reactors can provide a fairly steep wavefront of gate
95
SCR MANUAL
current together with a convenient means of control from a low level
DC or AC signal. This type of control is adaptable to feedback systems
and provides the additional advantage of multiple, electrically-isolated
inputs and outputs for more complex circuits.
4.13.1 Continuously Variable Control
A typical haH-wave magnetic amplifier type trigger circuit is shown
in Figure 4.26. The gate signal for triggering the SCR is obtained
from winding 3-4 of transformer T l • When the core of T:! is unsaturated, the winding 3-4 of T:! presents a high impedance to the gate
signal so that only a small voltage is developed across Ra. When
the core of T 2 saturates, the impedance of winding 3-4 of T 2 decreases
by several orders of magnitude so that a large voltage appears at the
gate of the SCR, causing it to trigger. Resistor R2 limits the gate current to the rated value and resistor Ra limits the gate voltage produced
by the magnetizing current of winding 3-4 of T 2 so that the SCR will
not trigger before the core of T 2 saturates. Diode CR 2 serves the dual
purpose of preventing a reverse voltage on the gate of the SCR and
preventing any reverse current through winding 3-4 which would produce an undesired reset of the core T 2.
--5r;------,
(SQUARE LOOP
COREl
LOAD
I
I
I
1
T
PHASE
CONTROL
ADJUST
FIGURE 4.26
TYPICAL HALF·WAVE MAGNETIC TRIGGER CIRCUIT
Control signals can be applied to either input 1 or input 2 or both.
Input 2 operates in the reset mode by controlling the reset voltage on
winding 1-2 of T!l during the negative half cycle. The setting of the
potentiometer Rl determines the amount of reset of the core during
the negative haH cycle, which in tum determines the phase angle of the
SCR conduction during the positive half cycle. Other control circuits,
such as a transistor amplifier stage, can be used in place of R l . Since
power is furnished by winding 5-6 of T" no auxiliary power supply is
needed. Input 1 operates in the MMF (magnetomotive force) made
by controlling the current through the winding 5-6 and the core flux
level, which in tum detHmines the trigger angle. The current for
input 1 must be obtained from an external power supply or from a
current generating type of transducer.
96
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
Additional output windings can be added to T!l for triggering
several SCR's in parallel or in series. Also, additional control windings
of the reset or MMF type can be added to T 2. Full wave and multiple
phase operation can be achieved by combining two or more half wave
circuits.
4.13.2 On-Off Magnetic Trigger Circuits
Magnetic trigger circuits designed for phase control applications
such as the one shown in Figure 4.26 require the use of saturable cores
which are large enough to allow the output winding to sustain the gate
voltage signal for a full half cycle without saturating. For simple on-off
control applications, the magnetic trigger circuits shown in Figure 4.27
permit the use of smaller and less expensive cores since the output
winding is not required to sustain the gate voltage signal for a full half
cycle. In addition, these circuits have the advantage of not requiring
the use of an auxiliary supply transformer.
+
1
SIGNAL
INPUT
AC
SUPPLY
AC
SUPPLY
j
j
(A) SHUNT CONFIGURATION
FIGURE 4.27
(8) SERIES CONFIGURATION
HALF-WAVE ON-OFF MAGNETIC TRIGGER CIRCUITS
In Figure 4.27(a), one winding of saturable transformer Tl is connected in shunt with the gate of SCR I . If T1 is unsaturated, the current
through R" R~ and CR I will flow into the gate of SCR I during the
first part of the positive half cycle and cauSe SCR I to turn on. If Tl is
saturated, the current through R" R~ and CR I will be diverted from
the gate by the low saturated impedance of the winding on T l . When
T 1 is saturated it can be reset, and the SCR can be made to trigger
by a positive voltage on the signal input. Capacitor C l provides filtering for the gate signal to prevent undesired triggering due to fast transients on the AC supply.
In Figure 4.27(b), one winding of saturable transformer T2 is connected in series with capacitor C~ and the gate of SCR!l. If T!l is
unsaturated the current through R;\ and CR!l will charge C!l during
the initial part of the positive half cycle. T:! will saturate after a few
degrees of the positive half cycle and permit a rapid discharge of C 2
into the gate of SC 2 , thus causing SCR!l to trigger. If T:! is initially saturated at the beginning of the positive half cycle, the winding of T 2 will
divert the current from C:! and prevent C:! from being charged. Resistor
97
SCR MANUAL
R4 prevents the voltage at the gate of SCR:.! produced by the current
through Rs from exceeding the maximum gate voltage that will not
trigger the SCR. When T 2 is saturated, it can be reset and the SCR
can be made to trigger by a positive voltage at the signal input.
The circuits of Figure 4.27 permit the SCR to perform thefunction of an AC contactor with an isolated DC control winding. Modifications of these circuits permit full wave operation with normally open,
normally closed or latching operation. The reader is referred to Chapter 8 for further discussion of static switching qircuits.
4.14 SEMICONDUCTOR TRIGGER·PULSE GENERATORS
The simple resistor and capacitor triggering circuits described in
Sections 4.12 and 4.13 depend heavily on the specific triggering characteristic of each SCR used. In addition, the power level in the control
circuit is high because the entire triggering current must flow through
the resistance. Furthermore, they do not readily lend themselves to
automatic, self-programmed, or feedback control systems.
Pulse triggering, on the other hand, can accommodate wide tolerances in triggering characteristics by overdriving the gate. The power
level in pulse control circuits may also be quite low since the required
triggering energy (IGT VGT t) can be stored slowly, then discharged
rapidly at the desired instant of triggering. The use of pulse triggering
enables small, low power, signal-type components and transducers to
control large, high-current thyristors, as shown in later chapters.
While there are a multitude of semiconductors and circuits which
can produce adequate triggering pulses, this chapter will consider only
those most adept at performing this function.
4.14.1 Basic Relaxation Oscillation Criteria
Most devices used to produce trigger pulses (such as: the unijunction transistor, diac trigger diode, the silicon unilateral. and bilateral
switches, programmable unijunction transistors, neon lamps, etc.)
operate by discharging a capacitor into the thyristor gate. They function in a basic relaxation oscillator circuit by means of a negative
resistance characteristic. Specifications for these devices usually include
the voltage and current required to achieve negative resistance when
approached from either the conducting or non-conducting states. (See
also Section 4.1.)
To relate these specifications to the criteria for oscillation, consider
the elementary relaxation oscillator circuit of Figure 4.28(a) using a
trigger device with voltage to switch V s, current to switch Is, holding
voltage VR , and holding current In. The device characteristic curve is
plotted in Figure 4.28(b), along with load lines representing Rl and R 2 •
If Rl is increased to the maximum value which will sustain oscillations,
we will find that its load line intersects the device curve at a point (1)
where the negative resistance slope of the device curve is equal to the
load line for R2 • This point (1) is very close to Is and VR, but not quite
the same since the specification of these values is made at the point
where the slope of the curve is vertical, representing zero dynamic
resistance.
98
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
When the triggering point (1) is reached, the operating point
transfers to point (2), discharging the capacitor with a peak pulse current, ip, and producing a peak pulse voltage ep, across the load resistor
R2 (which includes the thyristor gate impedance). The discharge of
the capacitor follows the device curve from point (2) to point (3),
where the negative resistance slope is once again tangential with the
R2 load line. The operation then transfers from point (3) to point (4),
the capacitor re-charges through Rl and the oscillation continues.
If Rl is changed to the minimum value which will sustain oscillation, its new load line will intersect the device curve at point (3). Any
smaller value will cause the device to remain conducting at some stable
lip
•
i
.p PULSE
~ OUTPUT
~------~~------~--~
(0 )
TRIGGER DEVICE CHARACTERISTIC
( b)
FIGURE 4.28
BASIC RELAXATION OSCILLATOR CIRCUIT AND CHARACTERISTICS
operating point between (2) and (3). Increasing Rl beyond the maximum oscillating value causes operation to cease at some point between
(1) and the origin.
A very important factor not apparent in Figure 4.28, and often
not specified for a device, is switching time, or rise time. A device
which slowly switches from point (1) to point (2) will never get there
since it is discharging the capacitor as it goes and will reach the device
curve somewhere between points (2) and (3). This switching time can
be a limiting factor if it is a significant fraction of the discharge timeconstant, R 2C.
The magnitude of pulse voltage, e l " and pulse current, ill' appearing at the load, resistor R2 in this circuit, is dependent upon the characteristic curve of the device and the relation between its switching
99
SCR MANUAL
time and the discharge time-constant, R2 C. For values of R 2C large
(> lOX) in comparison with the switching time of the device, the peak
pulse voltage, e p , is simply the difference between the switching voltage Vs and the conduction voltage drop VF' The peak pulse current
under this condition is found from the intersection of the R2 load line
and the characteristic curve.
When R 2 C is smaller, approaching the switching time, both e p
and ip are reduced by the effective device resistance during switching. As was shown in Section 4.8, reducing peak current, and
extending the pulse time accordingly, decreases the probability of triggering a thyristor.
Since the effect of switching time is not readily apparent from the
characteristic curve, devices intended for thyristor triggering generally
specify the peak pulse voltage across R2 (where the value of R2 is
chosen to represent typical gate impedance) when discharging a given
size capacitor typical for its application.
The following table shows the correlation of the parameter terminologies used in various switching devices with the points on the
general characteristic curve:
TABLE 4.1
Unilateral Devices
Terminology
On Figure 4.28
Bilateral Devices
UIT
SUS
PUT
SBS
ST4
Diac
Neon
Vs
V.
Vs
V.*
V.
Vs
V(BR)
V.
Is
I.
Vv
Iv
Is
1.*
I(BR)
Vv*
Iv*
Is
Vn
In
Is
Vn
IH
VOBl
Vo
ep
Vo
Vo
Vn
In
e.
ip
V.
ep
ip
'Determined Externally by Circuit
4.14.2 Unijunction Transistor
20
18
~
I.
I. /
l--- /
~EAK P~'NTS
I
~
+
12
)
v
VBB
I
BI
IE
>w
w
~
I
I
'B2
Nl-
VBB=~
'"
S
I
~
10
S
I--
-
I--
-
I-I
~
v vss=2iV
ffi
l::
1:']
•
-r--
Vee: IOV
.\~ ",v
BB'5V
~
V"t
/
EY
rOINTSI
182=0
10
EMITTE~
FIGURE 4.29
100
12
14
16
18
~
CURRENT - I E - MILLIAMPERES
GE 2N2646 UNIJUNCTION TRANSISTOR SYMBOL AND EMITTER
INPUT CHARACTERISTICS
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
The UJT has three terminals which are called the emitter (E),
base-one (Bl), and base-two (B 2). Between Bl and B2 the unijunction
has the characteristics of an ordinary resistance. This resistance is the
interbase resistance (RIlIl ) and at 25°e has values in the range from
4.7K to 9.1K.
The normal biasing conditions for a typical UJT are indicated in
Figure 4.29. If the emitter voltage, VE, is less than the emitter peak
point voltage, V1', the emitter will be reverse biased and only a small
reverse leakage current, lEO, will How. When VE is equal to VI' and
the emitter current, IE, is greater than the peak point current, I p , the
UJT will turn on. In the on condition, the resistance between the
emitter and base-one is very low and the emitter current will be limited
primarily by the series resistance of the emitter to base-one external
circuit.
The peak point voltage of the UJT varies in proportion to the interbase voltage, VBII , according to the equation:
(4.1)
The parameter 'YJ is called the intrinsic standoff ratio. The value of 'YJ
lies between 0.51 and 0.82, and the voltage VD , the equivalent emitter
diode voltage, is in the order of .5 volt at 25°e, depending on the
particular type of UJT. It is found that Vp decreases with temperature,
the temperature coefficient being about -3mvre for the 2N2646-47
(-2mvre for 2N489 series). The variation of the peak point voltage
with temperature may be ascribeu to the change in VD (also 'YJ for
2N2646-47 series). It is possible to compensate for this temperature
change by making use of the positive temperature coefficient of R BB .
If a resistor RIl2 is used in series with base-two as shown in Figure 4.30,
the temperature variation of Rlln will cause VIlll to increase with temperature. If RIl2 is chosen correctly, this increase in VIlll will compensate for the decrease in VI' in Equation 4.1. Over a temperature range
of -40 o e to lOOoe, Equation 4.3(a) gives an approximate value of
RB2 for the majority of 2N2646 and 2N2647 UJT's. Equation 4.3(b)
gives RB2 for the 2N489 MIL series, 2N1671A and B, and the 2N2160.
RIl?=~
YJV
(4.3a)
1
R 112
<=
OAORBIl
'YJVl
(1 - 'YJ)R Bl
+ -'--_.!..:.-..:o:.:..
YJ
(4.3b)
For a more detailed discussion of the characteristics of the various
types of UJT's the reader is referred to Reference 8. Quantitative data
and techniques for temperature compensation on an individual and
general basis in very high performance circuits over extreme temperature ranges are discussed in Reference 9.
101
SCR MANUAL
4.14.2.1 Basic UJT Pulse Trigger Circuit
VP
lZ1(--------- -
VE
TO
2v 0
------
------
VBI Gs,;:E
VBI~
FIGURE 4.30
BASIC UNIJUNCTION TRANSISTOR RELAXATION OSCILLATOR·TRIGGER
CIRCUIT WITH TYPICAL WAVEFORMS
The basic UJT trigger circuit used in applications with the SCR
is the simple relaxation oscillator shown in Figure 4.30. In this circuit,
the capacitor C 1 is charged through Rl until the emitter voltage
reaches Vp, at which time the UJT turns on and discharges C 1 through
Rm. When the emitter voltage reaches a value of about 2 volts, the
emitter ceases to conduct, the UJT turns off and the cycle is repeated.
The period of oscillation, T, is fairly independent of the supply voltage
and temperature, and is given by:
I I I
T = -f = Rl C 1 In -1-- = 2.3 RI C I loglo -1--."
-."
(4.4)
For an approximate nominal value of intrinsic standoff ratio of
." = 0.63, T = Rl C I .
The design conditions of the UJT triggering circuit are very broad.
In general, Rm is limited to a value below 100 ohms although values
up to 2 or 3K are possible in some applications. The resistor RI is limited to a value between 3K and 3 Meg. The lower limit on Rl is set
by the requirement that the load line formed by Rl and VI intersect
the emitter characteristic curve of Figure 4.29 to the left of the valley
point, otherwise the UJT in Figure 4.30 will not tum off. The upper
limit on RI is set by the requirement that the current flowing into the
emitter at the peak point must be greater than Ip for the UJT to tum on.
The recommended range of supply voltage VI is from 10 volts to 35
volts. This range is determined on the low end by the acceptable values
of signal amplitude and at the high end by the allowable power dissipation of the UJT.
If the pulse output (V m) of the circuit of Figure 4.30 is coupled
directly, or through series resistors, to the gates of the SCR's, the value
of Rm should be low enough to prevent the DC voltage at the gate
due to interbase current from exceeding the maximum voltage that
will not trigger the SCR's (see Figure 4.13) VGT (max) at the maximum
junction temperature at which the SCR's are expected to operate. To
meet this criterion, Rm should be chosen in accordance with the following inequality:
Rm VI
(4.5)
.)
R BB (mm
+ R m + R B2
o
3-,5
0
.'"
~ .4
:I:
U
/
~
.3
.2
V
A
,./"
/
IV
~
~~
~
-
CHARGE DELIVERED
BY AN EXPONENTIAL
PULSE: 80ma PEAK.
8/LSEC. TIME CONSTANT
~v
/
4
2
~
8
6
~
~
~
ffi
m
~
~
~
~
TIME (/LSEC.)
FIGURE 4.33
CHARGE TO TRIGGER AN EXPONENTIAL PULSE
The above technique provides a useful indication of SCR pulse
triggering requirements. However, it is important that a reasonable
safety factor be incorporated in the trigger circuit design, since there
are sizable variations in this characteristic.
Let Rs
39 ohms. Then since RR C T
8 ,...seconds, CT = 0.2 p.f.
The peak triggering current of 80 rna determines V p, namely:
=
VI'
=
= II> . RR + 1 V
= (80 rna) (39 n) + 1 V = 4.1
where 1 V is the approximate PUT on-state voltage
The computed value of 'Yf is:
__ VI) _ 4.1 ~ 1/3
Es
12
The timing pot RT can be found from Formula 4.7.
1
R'l'(max)
2.5 Meg
C T In 12 _ 4.1 f min
'Yf----~
=
RT(min)
(1)
=
1
12)
(
CT In 12 _ 4.1 f max
= 250K
The maximum valley current occurs at the maximum frequency when
R,l' is a minimum:
I. v(max)
E.
= -R--.=
T(mm)
12
250 rna
= 48 p.a
=
IV(min) of the 2N6027 is 70p.a for RG
10 K. Therefore, to find R}
and R 2 , the following two simultaneous equations must be solved:
107
SCR MANUAL
TIME CONSTANT MULTIPLIER
VS
GATE DIVIDER RATIO
FOR PUT OSCILLATOR
ASSUMMING vr' 5v, v v =.8V
4
V+
Jd"
t-----
I---
C
~~
R2
I
,
9
IOV
V'5V'
.,.
~
~
~
~
./
.4
~"
)~
2
.I
OSCILLATOR
I
PERIO~
• M X RC
I
J
.2
.4
•
•
~
.9
.8
GATE VOLTAGE DIVIDER RATIO RZ,ilRI +RZI
(a)
M
7
TIME CONSTANT MULTIPLIER
VS
GATE DIVIDER RATIO
FOR PUT T'MER
4
A5SUMMING vr=. 5V
v+
:;
"'
Q.
5
2
.i
C
z
~
8
,.
10
I
--
.........::
//
7
-/
~~
1
V
. / './". /
V""""': V
/ . V"/
4
/
Vh
o.I
V~V,~
R2
w
>=
IOV
]1J
0:
//
/'"
TIMING PERIOD': M x RC
I
V
.3
.4
.5
.6
.7
GATE VOLTAGE DivIDER RATIO R2j\RI+R21
.8
.9
(b)
FIGURE 4.34 EFFECT OF
108
""'(1, + L)
ON OSCIWTOR FREQUENCY
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
R _
G -
Rl R2
Rl + R2
R2
Rl
+ R2 -
Vp
Es
= 7J
The solutions for Rl and R2 are:
RG
Rl =-.7J
=
R-~
9~
1-7J
=
(4.9)
=
Since 7J
113, then Rl
30 K and R2
15 K.
If other frequency ranges had been desired, either different capacitors could have been switched in for C T or else R2 could have been
varied to achieve the same result.
Figure 4.34 shows the effect of the voltage divider ratio
R2/(Rl + R 2 ) on the period of oscillation.
4.14.4 Silicon Unilateral Switch (SUS)12
The SUS, such as the type 2N4987, is essentially a miniature SCR
haVing an anode gate (instead of the usual cathode gate) and a built-in
low-voltage avalanche diode between the gate and cathode. The symbol for the SUS and its equivalent circuit are shown in Figure 4.35.
Its anode-to-cathode electrical characteristic is shown in Figure 4.36
for no external connection to the gate terminal.
The SUS is usually used in the basic relaxation oscillator circuit
shown in Figure 4.28(a) and its characteristics follow the same criteria
for oscillation. The type 2N4987 has the following specifications:
Switching Voltage, V s .......... 6 to 10 volts
Switching Current, Is ........... 0.5 rna, maximum
Holding Voltage, VH . . . . . . . . . . . . Not specified (= 0.7 Vat 25°C)
Holding Current, I H . . . . . . . . . . . . 1.5 rna, maximum
Forward Voltage,
VF (at IF = 175 rna) .......... 1.5 volts
Reverse Voltage Rating, VR' . . . . . 30 volts
Peak Pulse Voltage, Vo .......... 3.5 volts minimum
The Peak Pulse Voltage, V0, specification is very important for
thyristor triggering applications since it is the only realistic figure-ofmerit that indicates the ability of the triggering device to transfer
charge from the capacitor to the thyristor gate. This voltage is measured
with the SUS operating in the circuit of Figure 4.28(a), where V1 = 15
10 K ohms, C
0.1 pi, and R2
20 ohms. The peak pulse
volts, Rl
voltage is measured across resistor R2. The magnitude of the pulse
voltage depends both upon the difference between V s and VF and upon
switching time, as explained in Section 4.14.1. The component values
used in the pulse test are adequate for triggering most thyristors.
The major difference in function between the SUS and the UJT
is that the SUS switches at a fixed voltage, determined by its internal
avalanche diode, rather than a fraction ("I) of another voltage. It should
also be noted that Is is much higher in the SUS than in the UJT, and
=
=
=
109
SCR MANUAL
is also very close to I H • These factors restrict the upper and lower limits
of frequency or time-delay which are practical with the SUS.
For synchronization, lock-out, or forced switching, bias or pulse
signals may be applied to the gate terminal of the SUS. For these
purposes, treat the SUS as an N-gate SCR.
4.14.5 Silicon Bilateral Switch (SBS)'2
The SBS, such as the type 2N4991, is essentially two identical SUS
structures arranged in inverse-parallel, as shown in Figures 4.37 and
4.38. Since its operates as a switch with both polarities of applied volttage, it is particularly useful for triggering the bidirectional· triode
thyristors (triacs) with alternate positive and negative gate pulses. This
operation is obtained by using an alternating voltage supply for VI of
Figure 4.28, rather than the DC supply shown.
Specifications for the SBS type 2N4991 are identical to those of
the SUS type 2N4987 with the exception of reverse voltage rating,
'rhich is not applicable to the SBS.
""~~ ""~~"~ _~Vf=R==-tt---=-t-+-.;-~='S~THODE
SYMBOL
\~CATHODE
VH VF
__ v
Vs
EQUIVALENT CIRCUIT
FIGURE 4.35
THE SILICON UNILATERAL SWITCH (SUS)
--- -,
I
r--I
I
FIGURE 4.36
SUS CHARACTERISTIC CURVE
I
't.,
GATE
I
GATEo--h~
I
I
I
I
~
.
L
2
SYMBOL
-
_J
EQUIVALENT CI ReUIT
FIGURE 4.37
THE SILlCON81LATERAL SWITCH (S8S)
FIGURE 4.38
S8S CHARACTERISTIC CURVE
4.14.6 Bilateral Trigger Diode (Diac)
The diac, such as the type ST2 is essentially a transistor structure,
Figure 4.39, wh,ich exhibits a negative resistance characteristic above
a given switching current I(BRl' The characteristic curve of Figure 4.40
shows that this negative resistance. region extends over the full operatingrange of currents above I(BR) hence the concept of a holding
current IH does not apply.
110
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
The diac is used in the simple relaxation oscillator circuit of Figure
4.28, and the criteria for oscillation are the same. For alternating pulse
output, the supply voltage for the oscillator circuit, Vh may be an
alternating voltage.
FIGURE 4.39
SYMBOL OF BILATERAL TRIGGER DIDDE (DIAC)
FIGURE 4.40
DIAC CHARACTERISTIC CURVE
The type ST2 diac has the following specifications:
V (BR) . . . . . . . . . . . . . . . . . . . . 28 to 36 volts
I(BR) . . . . . . . . . . . . . . . . . . . . . 200 /Lamp (maximum)
ep . . . • • • . • . . . . . . . . . . . . • • . 3 volts (minimum)
The peak pulse voltage, ep , is measured under the same conditions
20 ohms; C
0.1 microused with the SUS and SBS, namely: R2
farad. Since the ST2 is primarily used to trigger triacs, this minimum
value of e p has been established to ensure proper triggering of all G-E
triacs, assuming, of course, the proper conditions of supply voltage and
load impedance in the power circuit of the triac.
=
=
4.14.7 Asymmetrical AC Trigger Switch (ST4)
The ST4 is an integrated triac trigger circuit that provides wide
range, hysteresis-free, phase control of voltage. This performance is
possible with a minimum number of circuit components and at very
low cost (see Chapter 7 for circuit details).
The equivalent circuit of Figure 4.41 reveals that the ST4 behaves
like a zener diode in series with an SBS. The zener diode provides the
asymmetry since now switching voltage V81 has been increased by the
avalanche voltage of the zener.
FIGURE 4.41
+. *-
SYMBOL OF ASYMMETRICAL AC TRIGGER SWITCH (ST4)
AND EQUIVALENT CIRCUIT
111
SCR MANUAL
FIGURE 4.42
ST4 CHARACTERISTIC CURVE
The ST4 has the following specifications:
14-18 volts
Switching Voltage: V S1
VS2
7-9 volts
Switching Current: ISh IS2 80 p;J. (25°C)
ISb IS2 160 /La (-55°C)
On-State Voltages: VF1
7-10 volts
VF2
1.6 volts (max)
Peak Pulse Voltage, V0
3.5 volts minimum
4.14.8 Other Trigger Devices
Several other unilateral and bilateral switching devices exist,
having characteristics similar to those discussed above. In general, all
operate as relaxation oscillators and are subject to the same criteria for
oscillation. If the peak pulse voltage (or current) output is not specified,
then the maximum switching time must be known. Otherwise, the trigger circuit must be over-designed by a factor depending upon the
uncertainty of the unknowns.
Where a large demand exists for the same type triggering source,
specialized integrated circuits could be and have been designed to
meet the need. The GEL300 is a monolithic integrated triggering
circuit that features "zero-voltage" switching to minimize RFI. Its operation and use is covered in Chapter 11. Similarly the GEL301 IC is
used in phase control circuits (see Chapter 9).
Another method of triggering that is coming into its own employs
light sensitive or light activated devices. This method offers speed along
with incomparable electrical isolation. The reader is referred to Chapter 14 and Reference 16 for additional information.
4.14.9 Summary of Semiconductor Trigger Devices
The table below summarizes the electrical characteristics of the
widely used triggering devices.
112
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
'·1
CLASS
CHARACTERISTICS
a
_.. ~
~~
WT
Unijunctilln
Transidor
Unijunction
Trallliltar
E
SUS
u_
......
E.
6
.,_
SCS
.....
......,
S8S
.....
Silicon
i
I
DIAC
.n
Assym.triul
AC'_
hlkh(Sl'4)
mil
2N489A
l-
2N2417A
2N1671C
Vo_
'N....
2N6027
2Nfi028
...... ........
......
.........,
......
Il>li
Il>lIl
IN....
'N_
.,
"2['
ii+ii2
7·tv
"2[,
(4Owmall
Il>li
2N4993
~. M
+~
6-l0y
1.5-h
7.S-B.h
iij+i2-
3N. .
ST'
.n
....
PEl~~~
CURRENT
Ivlminl
TURN ON
SPeCIFICATION
TIME
NUMBER
....---..--
12 ...
0 ..
"'"
'"
'"
0"
Allow.2,..
7<1"
AIIow • . lSp.
(A functiOfl of
tAfunctionof
R,.A21
R"A21
""
...,....."
,.....
300"
'-2$1_
Typ
.0 _
1.0
nIB
.7S_
-
......."'''..........
....
80. .
"',-
On
1.5 ...
lOrna
Afunctianof
", and R2
TON
VALLey
CURRENT
'"
2N2647
~~
L
F....
,_....
Fraction
of
fJ;'
--, K:
~
7
Il>li
2N489B 2N2417B
2N1671A 5G515
2N1871B &0516
.1..1..
[6 hi
Silicon
PEA:JoINT
VOLTAGE
~RI
PUT
SiIiI;on
MAJOR TVPES
BASIC CIRCUIT
1.0$1_
Mn
1.5,,_
Mn
(A function of
00..
85.25,65.28
65.27, 6&.28
85.27,65.28
85.26,86.26
65."
R, andR2i
,Il>IO ,
e.1Ov
7.S-lv
600"
120,..
1.5 rna
2N4992
28 v-36 v
200 ..
Very high
.lima
1.0,,_
85.30,85.31
Mn
1Ii.32
"_
176.30
Typ
14.18V
l-9V
...
~.
1~_
n5.32
~.
E
TABLE 4.2
4.15 NEON GLOW LAMPS AS TRIGGER DEVICES
The low price of neon glow lamps has led many to consider their
use for triggering thyristors. The characteristics of the glow lamp are
quite similar, but for magnitude, to those of the diac. The switching
voltage is generally on the order of 90 volts and the switching current
is extremely small (below 1 p,a). However, the switching time is large
in comparison with semiconductor devices, and the peak pulse voltage
is usually not specified.
The G-E type 5AH is an isotope-stabilized neon glow lamp now
being used in many low-cost SCR control circuits. The 5AH lamp has
the following specifications:
V s ................... 60 to 100 volts
Is .... . . . . . . . . . . . . . . . .. Not specified
VF . . . . . . . . . . . . . . . . . . . Approx. 60 volts at 5 ma
VH . . . . . . . . . . . . . . . . . . . Not specified
IH .................... Not specified
ip .................... 25 ma (minimum)
113
SCR MANUAL
The peak pulse current, il" is measured in a 20 ohm resistor when
discharging a 0.1 p.f capacitor. The minimum peak pulse voltage is,
therefore, 0.5 volts under this condition. The specification also includes
an indication of the operating life of the lamp: 5000 hours operation,
on the average, at 5 rna DC results in a 5 volt change in Vs or VF' This
has not been correlated to hours operation in a relaxation oscillator at
120 Hz.
Glow lamps are useful for thyristor triggering under the following
conditions:
(a) Thyristor IGT on the order of 10 rna or less
(b) Wide tolerance in Vs is acceptable
(c) Minimum pulse voltage measured in sample lot several times
minimum required to trigger the thyristor
(d) Change in Vs and pulse output with operating time is
acceptable
(e) Cost of primary importance
(f) 5% loss in RMS voltage at full power tolerable
4.15.1 Neon Lamp Trigger Circuits
Neon lamp SCR phase-controlled trigger circuits have the promise
of combining the low cost of the RC diode circuit with improved performance. In addition, the possibility exists in such a relatively simple
yet high impedance circuit to exercise control over the charging rate
of the trigger capacitor with suitable devices responsive to light, heat,
pressure, etc.
Figure 4.43 shows a half wave AC phase-controlled circuit using
a 5AH as the trigger for a two terminal system. The 5AH will trigger
when the voltage across the two 0.1 MFD capacitors reaches the breakdown voltage of the lamp. Control can be obtained full off to 95% of
the half wave RMS output voltage. Full power can be obtained with
the addition of the switch across the SCR.
3K
GE
C228
SWITCH
FOR FULL
POWER
0.1 lIFO
FIGURE 4.43
HAlf WAVE/TWO TERMINAL
Figure 4.44 is a transformer coupled full-wave AC phase-controlled
circuit using a 5AH as the trigger for a two terminal system. The 5AH
will perform the same as in the half-wave circuit but the pulse transformer will allow the SCR's to alternate in firing. The resistor Rand
the pulse transformer should be chosen to give pmper shape of the
pulse to the gate of the SCR. Some loss of load voltage will occur but
will amount to only about 5% in terms of total RMS output voltage.
114
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
A
B
SCR'S
GE
C-22
x
y
FIGURE 4.44
FULL WAVE TRANSFORMER COUPLED/TWO TERMINAL
4.16 PULSE TRANSFORMERS
Pulse transformer are often used to couple a trigger-pulse generator to a thyristor in order to obtain electrical isolation between the two
circuits. There are many vendors of pulse transformers suitable for this
purpose. Although several specific model numbers are shown on circuit
diagrams in this manual, it is not our purpose nor intent to serve as a
testing or approval function.
The transformers usually used for thyristor control are either 1: 1,
two-winding, or 1: 1: 1 three-winding types. As shown in Figure 4.45,
the transformer may be connected directly between gate and cathode,
or may have a series resistor R to either reduce the SCR holding current or to balance gate currents in a three-winding transformer connected to two SCR's, or may have a series diode D to prevent reverse
gate current in the case of ringing or reversal of the pulse transformer
output voltage. The diode also reduces holding current of the SCR.
In some cases where high noise levels are present, it may be necessary
to load the secondary of the transformer with a resistor to prevent false
triggering.
o
TRIGGER PULSE GENERATOR
T.PG.
~I
0
•
1·1 PULSE
TRANSFORMER
FIGURE 4.45
BASIC PULSE TRANSFORMER COUPLING
Figure 4.46 shows several ways of using a transformer to drive
an inverse-parallel pair of SCR's. Full isolation is provided by the three115
SCR MANUAL
winding transformer in Figure 4,46(a). Where such isolation is not
required, a two-winding transformer may be used either in a series
mode, Figure4,46(b), or a parallel mode, Figure 4,46(c). In any case,
the pulse generator must supply enough energy to trigger both SCR's,
and the pulse transformer (plus any additional balancing resistors) must
supply sufficient gate current to both SCR's under worst-case conditions
of unbalanced gate impedances.
UNILATERAL
T.P.G.
SCR2
IQ)
1'1
PULSE
TRANSFORMER
•
UNILATERAL
TP'G.
+
TPG.
Ib)
FIGURE 4.46
Ie)
PULSE TRANSFORMER CONNECTIONS FOR TWO SCR'S
The prime requirement of a trigger pulse transformer is one of
efficiency. The simplest test is to use the desired trigger pulse generator
to drive a 20 ohm resistor alone and then drive the same resistor
through the pulse transformer. If the pulse waveforms across the
resistor are the same under both conditions, the transformer is perfect.
Some loss is to be expected, however, and must be compensated by
increased drive from the generator.
Some of the transformer design factors to be considered are:
(a) Primary magnetizing inductance should be high enough so
that magnetizing current is low, in comparison with pulse current,
during the pulse time.
(b) Since most pulse generators are unilateral, core saturation
must be avoided.
(c) Coupling between primary and secondary should be tight, for
single-SCR control, or may have specified leakage reactance to assist
in balancing currents for multiple-SCR control.
(d) Insulation between windings must be adequate for the application, including transients.
(e) Interwinding capacitance is usually insignificant but may be
a path for undesirable stray signals at high frequencies.
116
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
4.17 SYNCHRONIZATION METHODS
In the basic trigger circuit of Figure 4.47, the UJT can be triggered at any intermediate part of the cycle by reducing either the
interbase voltage alone or the supply voltage, V l' This results in an
equivalent decrease in VI' in accordance with Equation 4.1 (or 4.3)
and causes the UJT to trigger if VI' drops below the instantaneous
value of Vg. Thus, the base-two terminal or the main supply 'voltage
can be used to synchronize the basic trigger circuit. Figure 4.47 illustrates the use of a negative synchronizing pulse at base-two.
+<>-.-------,
FIGURE 4.47
PULSE SYNCHRaNIZATION OF UJT RELAXATION OSCILLATOR
Two methods of achieving synchronization with the AC line are
illustrated in Figure 4.48. A full wave rectified signal obtained from
a rectifier bridge or a similar source is used to supply both power and a
synchronizing signal to the trigger circuit. Zener diode CR 1 is used to
clip and regulate the peaks of the AC as indicated in Figures 4.48 (a)
and (b) .
.a:n..
J:Dl
,/,/
.£r:£:L
,/
J:Dl
I
CR2
I
C2
AC
FULL-WAVE
T
OUTPUT
OUTPUT
TO SCR
.--+-+TO SCI!
GATE
BATE
RBI
(A)
(B)
FIGURE 4.48
CIRCUITS FOR SYNCHRONIZATION TO AC LINE
At the end of each half-cycle the voltage at base-two of Ql will
drop to zero, causing Ql to trigger. The capacitors C 1 are thus discharged at the beginning of each half cycle and the trigger circuits
are thus synchronized with the line. In Figure 4.48(a) a pulse is produced at the output at the end of each half cycle which can cause the
SCR to trigger and produce a small current in the load. If this is
undesirable, a second UJT can be used for discharging the capacitor
117
SCR MANUAL
at the end of the haH cycle as illustrated in Figure 4.48(b). Diode CRl
and capacitor C 2 are used to supply a constant DC voltage to Q2' The
voltage across Ql will drop to zero each half-cycle causing C l to be
discharged through Ql rather than through the load RBl . The UJT's
should be chosen so that Ql has a higher standoff ratio than Q2'
Synchronization of a PUT circuit is exactly analogous to the UJT
since their operation is so similar.
4.18 TRIGGER CIRCUITS FOR INVERTERS
Inverter circuits usually require trigger pulses delivered alternately to two SCR's. There are many ways and types of circuits to perfonn this function, several of which are mentioned below.
4.18.1 Transistorized Flip-Flops
The transistor flip-flopS is a very fundamental and useful circuit
for driving SCR or triac gates. The transistor may drive the gates
directly, as described in Section 8.9.2, through a transfonner or through
a pulse shaper as shown in Section 4.19. The transfonner must be
designed to avoid saturation at the lowest operating frequency and
highest supply voltage. The flip-flop may be driven by a UJT or a PUT
relaxation oscillator for precise timing or may be connected as a freerunning multivibrator.
TO SCR
GATE
(A)
TO SCR
GATE
*MUST BE HEAT SUNK
-=
I KHz OSCILLATION FOR COMPONENTS SHOWN
22V
33K
TO SCR
GATE
(B)
82K
TO SCR
GATE
*MUST BE HEAT SUNK.
500 Hz OSCILLATION AS SHOWN.
FIGURE 4.49
118
82K
FLlP·FLOP TRIGGER CIRCUITS FOR TWO INVERTER SCR'S
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
Figure 4.49 shows two approaches which provides alternate output pulses such as required by many inverter circuits. Alternate output
pulses are obtained by cross coupling two relaxation oscillator circuits
by capacitor C 1 • Frequency is trimmed by potentiometer RI and symmetry is trimmed by R 2 • Both circuits offer the same rise times and
have an upper frequency limit of 20 kHz but the circuit with the PUT
does possess greater versatility and higher output voltages. The oscillation frequency of the latter can be varied either by changing the capacitors or by varying the gate bias on the PUT.
4.18.2 PUT Flip-Flop Trigger Circuit
This flip-flop circuit consists of two relaxation oscillator circuits
coupled together as shown in Figure 4.50. When one of the two trigger
devices is in the "on state," the other is always in the "off state." Turning on one device will instantaneously produce a negative voltage on
the other due to the presence of capacitor CT. This will shift it to the
"off state." The frequency is adjusted by RI and the symmetry is
trimmed by R 2 • Outputs VI and V2 may be coupled to additional stages
of amplification before coupling to the gate.
\Is
VI
470
RI
IK
470
100
FIGURE 4.50
v2
Vlh b
v2b b ~
FlIP·FLOP TRIGGER CIRCUIT
4.19 PULSE AMPLIFICATION AND SHAPING
Consideration of SCR trigger requirements may reveal that the
output of a pulse generator is not of sufficient amplitude and/or its
output rise time is too slow. With little additional expense, the output
can be bolstered to meet the stringent gate requirements of SCR's working'at·high frequencies and high di/dt.
Figure 4.51 shows several gate amplifier circuits. Circuit (a)
utilizes a transistor amplifier which is saturated during the duration of
the relaxation oscillator pulse. This allows C I to discharge into the SCR.
The availability of SCR's with highly sensitive gates permits use
of these devices to trigger higher rated SCR's as shown in Figure
4.51 (b). Here, for example, a C5B as SCR I requires less than 200
microamperes of gate signal to trigger. Current then flows through R2 ,
SCRb and into the gate of SCR 2 • When the current reaches the triggering requirements of SCR2 , this device turns on and shunts the main
power away from SCRI . In addition to providing a means of triggering
high current SCR's by low level signals from high impedance sources,
this type of triggering yields positive triggering from pulsed gate signals even with highly inductive loads due to the much lower latching
current requirements of the C5 in comparison with the higher rated
119
SCR MANUAL
SCR's. With SCR1 latched into conduction, the gate of SCR2 is driven
by a trigger signal which is maintained until SCR2 is forced into conduction. R 2 limits the current through SCR 1 to a value within its rating.
SCR 1 must meet the same voltage requirements as SCR2 • However, its
current duty is generally of a pulsed nature, and hence negligible. Several types of SCR's such as the C398 and C158 have amplifying gates,
in which the predriver SCR's are internal to the device as shown in
Figure 4.5l(c).
.
FROI! RELAXATION
OSCILLAlOR
'
:iJ
QI
C-I
I
IOKVA
R2
5CH 2
G£ CtaOS
120 VAC
r- --I
S~:::L
I
:
SOURCE:
'-
(0) TRANSISTOR PULSE AMPUFIER
I
:
IK:
__ ...J
Ib) USE OF LOW CURRENT SCR AS A GATE
SIGNAL AMPLIFIER
Ie) AMPLIFYING GATE SCR
FIGURE 4.51
TRIGGER PULSE AMPLIFIER CIRCUITS
Predrivers are particularly useful when the trigger voltage must
be maintained for the entire conduction period. Under these conditions,
the power dissipation in the gate of the power device may be excessive.
Figure 4.52 provides a technique of maintaining trigger drive during
the conduction period in the form of a pulse train and thus r~ducing
the average gate dissipation. The transistor multivibrator, provides
alternate driving voltages to the two unijunction transistor oscillators.
The outputs of these oscillators provide the alternating pulse train
sequence as required for inverter circuits.
330
IK
18K
18K
330
IK
10K
+
: 28VDC
TO SCR
~
GATE
JIIIIIIIU~IIIII[HZ
~
50Hz
Q( a Q2 - 2N3416
Q3
Q4 - 2N2647
NOTE: FOR SOURCE VOLTAGES
LESS THAN 25 VDC
USE 2N3414 FOR QI
Q2
a
FIGURE 4.52 TRIGGER CIRCUIT PROVIDING TRAIN OF PULSES
120
a
GATE TRIGGER CHARACTERISTICS, RATINGS, AND METHODS
For some high current switching applications, it is desirable to
trigger with a fast rise-time pulse. A slow rising pulse may be sharpened by use of the circuit in Figure 4.53. When a pulse appears at the
input of this circuit, the diode Dl will conduct, charging the capacitor.
The forward drop across the diode will assure a positive gate to anode
voltage on the PUT and will prevent it from switching. When the
capacitor charges to the peak voltage of the pulse, the diode will
become reverse biased and the PUT will switch on. The consequent
pulse delivered to the SCR will have a rise time of 50 to 100 nanoseconds determined by the PUT turn-on characteristics.
FIGURE 4.53
PULSE SHARPENER USING A PUT
It has been noted that for fast rising current loads, an SCR may
require a fast rising high level rectangular pulse to assure triggering.
Rectangular pulses can be shaped by the use of reactive pulse forming
networks or by blocking oscillators. However, these circuits are relatively costly and large. Figure 4.54 shows a circuit which will generate
rectangular pulses of 10 p.Seconds pulse width at repetition rates up to
20 kHz and does not require any inductive elements. With a 20 volt
amplitude and 20 ohm source impedance, this circuit should adequately
trigger most SCR's even under the most stringent di/dt conditions.
The UJT operates as a conventional relaxation oscUlator whose frequency may be controlled by any of the techniques that were previously mentioned. The UJT output pulses drive a four transistor amplifier
circuit which improves the rise time and extends the pulse width to
approximately 10 ,useconds.
1.0
TO SCR GATE
0, - 2N2647
°2.~-2N34'4
03 - 2N5365
05- D43C2
FIGURE 4.54
HIGH di/dt TRIGGER CIRCUIT
121
SCR MANUAL
Depending on the nature of the control input signal other types
of SCR's can be considered for triggering larger SCR's. The LASCR
(Chapter 14) can be used where direct triggering by light is required.
Also, the LASCR in conjunction with a suitable light source provides
a simple way in which to obtain electrical isolation in SCR control
circuitry.
REFERENCES
1. "Using Low Current SCR,s," D. R. Grafham, General Electric
Company, Auburn, New York, Application Note 200.19.*
2. "Semiconductor Controlled Rectifiers," F. E. Gentry et aI., Prentice Hall, Englewood Cliffs, New Jersey, 1964, Chapter 5.
3. "An All Solid-State Phase Controlled Rectifier System," F. W.
Gutzwiller, AlEE Paper CP 59-217, American Institute of Electrical Engineers, New York, N. Y., 1959.
4. "The Unijunction Transistor: Characteristics and Applications,"
T. P. Sylvan, General Electric Company, Syracuse, N. Y., Application Note 90.10.*
5. Glow Lamp Manual and Miniature Lamp Bulletin 3-3474, General Electric Company, Nela Park, Cleveland, Ohio, 1963.
6. "Silicon Controlled Switches," R. A. Stasior, General Electric Company, Syracuse, N. Y., Application Note 90.16.*
7. "Transistors and Active Elements," J. C. Linville and J. F. Gibbons,
McGraw-Hill Co., New York, 1961.
8. Transistor Manual, 7th Edition, "Unijunction Transistor Circuits,"
General Electric Company, Syracuse, N. Y., 1964, Publication
450.37.*
9. "Unijunction Temperature Compensation," D. V. Jones, General
Electric Company, Syracuse, N. Y., Application Note 90.12.*
10. "Electronic Circuit Theory," H. J. Zimmerman and S. J. Mason,
John Wiley and Sons, New York, N. Y., 1960, pp. 467-476.
11. "A Handbook of Selected Semiconductor Circuits," Seymour
Schwartz, Editor, Bureau of Ships, Department of the Navy, Publication NAVSHIPS 93484, pp. 6-18 to 6-25.
12. "Using the Silicon Unilateral and Bilateral Switches," R. Muth,
General Electric Company, Syracuse, N. Y., Application Note
90.57.*
13. "A 15KC, DC to DC Converter," J. A. Pirraglia and R. Rando,
IEEE Conference Record of the Industrial Static Power Conversion Conference, No. 34C20, The Institute of Electrical and Electronic Engineers, New York, N. Y., 1965.
14. "Design of Triggering Circuits for Power SCR's," J. M. Reschovsky,
General Electric Company, Syracuse, N. Y., Application Note
200.54.*
15. "The D13T - A Programmable Unijunction Transistor Types
2N6027 and 2N6028," W. R. Spoilard, Jr., General Electric Company, Syracuse, N. Y., Application Note 90.70.*
16. "The Light Activated SCR," E. K. Howell, General Electric Company, Syracuse, N. Y., Application Note 200.34.*
·Refer to Chapter 23 for availability and ord~ring information.
122
5
DYNAMIC CHARACTERISTICS OF SCR'S
DYNAMIC CHARACTERISTICS OF SCR'S
Dynamic characteristics of an SCR refer to the SCR behavior
during switching intervals. This may be switching of the SCR under
consideration (either turn-on or tum-off) or switching elsewhere in the
circuit (resulting in high dv/dt applied to the SCR). Although this
typically represents a minor percentage of the time period, it often
demands major consideration by both the manufacturer and user of
the SCR.
During the turn-off, or commutation, interval the SCR characteristics to be considered are tum-off time, reapplied dv I dt, and reverse
recovered charge. Each of these dynamic characteristics are dependent
upon the set of operating conditions of the specific circuit.
During the tum-on interval the dynamic condition to be considered by the circuit designer is the rate of rise of forward current, di/dt.
Switching losses during both turn-on and tum-off may be of concern
to the equipment designer.
Another dynamic condition, a fast-rising forward voltage, can
result in the SCR being switched from the blocking state to the on-state.
5.1 SCR TURN·OFF TIME, tq
If forward voltage is applied to a undirectional thyristor (hereafter
referred to as "SCR") too soon after anode current ceases to flow, the
SCR will go into the conduction state again. It is necessary to wait for
a definite interval of time after cessation of current flow before forward
voltage can be reapplied. Chapter 1 describes the physical reasons for
this required interval. For turn-off-time as it affects bi-directional
thyristors, see Chapter 7.
To measure the required interval, the SCR is operated with current and voltage waveforms shown in Figure 5.1. The interval between
ta and ts is then decreased until the point is found when the SCR will
just support reapplied forward voltage.
This interval is not a constant, but is a function of several parameters. Thus, the minimum time ta to ts will increase with:
1. An increase in junction temperature.
2. An increase in forward current amplitude (t1 to t 2).
3. An increase in the rate of decay of forward current (t2 to t a).
4. A decrease in peak reverse current (t4 ).
5. A decrease in reverse voltage (til to t7).
6. An increase in the rate of reapplication of forward blocking
voltage (ts to t9).
7. An increase in forward blocking voltage (t9 to tlO).
8. An increase in external gate impedance.
9. A more positive gate bias voltage.
123
SCR MANUAL
\
~(I
e
----l-
I'!, ----- I
\I
I,
FIGURE 5.1
f2
I
fa
SCR WAVEFORM FOR
TURN~FF
l
17
Ie'
1'0
TIME MEASUREMENTS
5.1.1 SCR Turn·Off Time Definitions
Circuit commutated tum-off time of an SCR, tq, is the time interval between the instant when the anode current has decreased to zero,
and the instant when the SCR has regained some defined forward
blocking voltage capability.
As mentioned in Section 5.1, tq is not a constant, but is dependent
upon the test conditions under which it is measured. One of these test
conditions, forward current (IT)' is shown in Figure 5.2 for a narrow
pulse of current and in Figure 5.3 for conventional circuit tum-off
time. Refer to Chapter 2 for definition of terms used in the figures.
f Ii!
~
!Z
~
tie
o+---+,--+r-----
~ I~
~
4
!
..
:
-.tIp
r,,,
~
- IRM
I
I
,-lRY
I
I
,
I
"'-tq~
,
I
t - - IqIPULSE)-..
I
I
I
I
,,
,,
I
v"'"'
I
,,
I
I
I
FIGURE 5.2 PULSE CIRCUIT COMMUTATED
TURN-GFF TIME
124
FIGURE 5.3
CONVENTIONAL CIRCUIT COM.
MUTATED TURN-GFF TIME
DYNAMIC CHARACTERISTICS OF SCR'S
5.1.2 Typical Variation of Turn-Off Time
The extent to which the parameters in Section 5.1 affect turn-off
time is dependent on both the parameter being considered and the
device design. Turn-off time performance trade-off curves are used to
determine which parameters are of significance, depending on the
particular set of circuit operating conditions. Figure 5.4 for example
shows a typical curve of change in SCR turn-off time with junction
temperature for a specific SCR. Figure 5.5 shows the dependence of
tq on forward current, for rectangular current pulses for that same SCR.
12
III
~
o
U
ILl
III
o
a:
10
8
~
~
::;:
I
6
/
...............
-
ILl
::;:
i=
......
o
4
V
/
V
GE-CI41 SCR
2
o
o
40
20
60
80
100
120
140
JUNCTION TEMPERATURE - °C
FIGURE 5.4 VARIATION OF TURN-OFF TIME WITH JUNCTION TEMPERATURE
o
ILl
III
NOTES'
(1) RECTANGULAR CURRENT PULSES, 50
MICROSECONO MINIMUM" DURATION
(21 MAXIMUM CASE TEMPERATURE = 120°C
(3) RATE OF RISE AND FALL OF CURRENT
LESS THAN 10 AMPERES PER P.SEC
(4) FREOUENCY 50 TO 400 Hz
(5) MAXIMUM dv/dt = 200 VOLTS PER P.SEC
::I..
~
20
j::
......
o
z
18
a:
16
a
14
:::l
IILl
l-
e(
,....
:::l
::;:
::;:
o
o
l-
S
u
a:
u
GE-1C141
~
12
10
8
I--
V
.,/
--
~
o
10
20
30
PEAK
40
50
60
70
80
90
100
FORWARD CURRENT - AMPERES
FIGURE 5.5 VARIATION OF TURN-OFF TIME WITH PEAK FORWARD CURRENT
FOR A HIGH SPEED SCR
125
SCR MANUAL
Typical variation of turn-off time with applied reverse voltage is
seen in Figure 5.6 for the C158 SCR.
+SO
OJ
::E
......~
+40
0
I
Z
0:
'"
f- + 20
OJ
!:
.....
~f
OJ
0
0
Z
>
l:
0
u
fZ
OJ
u -20
0:
~
11-
40
-so
I
REVERSE VOLTAGE-VR IN VOLTS - - -
FIGURE 5.6
TYPICAL VARIATION OF TURN·OFF TIME WITH APPLIED REVERSE
VOLTAGE DURING THE TURN·OFF INTERVAL
Typical dependence of turn-off time on another test condition,
reapplied dv/dt, is shown in Figure 5.7. Specified turn-off time for the
C158 is with a reapplied dv/dt test condition of 200 volts per microsecond.
+60 r-----.---lr-.]-.rT-.rrr-----.---r---------~
::E
:;:j
~
.:.
g
!oJ
i= +40
:
--'
+20
0
2
V ORMrJ
I
dv 1011-
I-
i
1\ .J.JI-vR
III
~
ffifh~I:-VTM
~
0
~
-20
~ I ~7.5
u
u
0
~
~
Iq. 4O,.sec OR LESS - ,
I
I I I
I....f'
I
1
el5e
AMPS/Jl.SEC
V
1-
ffi
-+-+-1+1+------+--+-----------1
L
V
CONDITIONS:
I
-
_1-_________
ON-STATE CURRENT ( I Tl= 150 AMPS
OFF-STATE VOLTAGE (VDRM1= RATED
~ -40 ~____-l-__-I----tREP. PEAK REV. VOLTAGE (VRRM1S RA~
REVERSE VOLTAGE (VR1=50 VOLTS
~
CASE TEMPERATURE (Tc 1=125"C
a:
~~\l'i~.fs ~!'s'Ec0F FIRWA~D CURRENT-60 L-____L-~__L-~~~~____L-~________~
10
20
30 40 5060 80 100
200
REAPPLIED dV/dt-VOLTS/" SEC __
FIGURE 5.7 TYPICAL VARIATION OF SCR TURN·llFF TIME WITH
REAPPlIED OVIOT
126
DYNAMIC CHARACTERISTICS OF SCR'S
5.1.3 Circuit Turn-off Time (te)
Circuit turn-off time is the turn-off time that the circuit presents
to the SCR.
The circuit turn-off time (4) must always be greater than the turnoff time of the SCR (tq ); otherwise the SCR may revert to the on-state.
The turn-off times of general purpose phase control SCR's are
usually given as typical values if at all. Wide deviations from the
typical values can occur. In those circuits where turn-off time is a
critical characteristic, it is necessary for the circuit designer to have
control over the maximum value of SCR turn-off time. For this reason
General Electric offers a range of SCR's with guaranteed maximum
turn-off times under specified standard conditions of waveform and
temperature. These SCR's are designed specifically for inverter applications where the need for good dynamic capabilities is essential.
5.1.4 Feedback Diode
Many inverter circuits require the use of a feedback diode placed
in inverse parallel with the SCR. The diode is needed to carry reactive
energy to the supply during some portion of the operating period. This
diode is a disadvantage from an SCR turn-off standpoint. The reverse
voltage of the SCR during the commutation interval is limited to the
diode forward voltage thereby adversely affecting SCR turn-off time as
discussed earlier in this chapter.
The feedback diode, when needed, should be placed close to the
SCR in order to minimize inductance in the diode path. As shown in
the sketch, this inductance undesirably shortens the circuit turn-off
time because of the voltage induced in the inductance due to the changing current (V = L di/dt).
(a) Diode adjoining SCR,
No Inductance
(b) Inductance in Diode Path
5.2 TURN-OFF METHODS
The gate has no control over the SCR once anode-to-cathode
current exceeds latching current. External measures therefore have to
be applied to stop the How of current. There are two basic methods
available for commutation, as the turn-off process is called.
127
SCR MANUAL
5.2.1 Current Interruption
The current through the SCR may be interrupted by means of a
switch in either of two circuit locations. The switch must be operated
for the required tum-off time. Note that the operation of the switch will
cause the SCR to see high values of dv/dt.In Figure 5.8(a) when the
switch is closed, or in 5.8(b) when the switch is opened, the SCR is
susceptible to false tum-on due to high dv I dt.
As it is seldom that a mechanical switch is suitable for commutation, various static switching circuits have been developed for this
purpose. It should be noted that SCR's are generally not characterized
for this mode of commutation.
1
(a)
(b)
FIGURE 5.8
COMMUTATION BY CURRENT INTERRUPTION
5.2.2 Forced Commutation
When the above methods of current interruption are not acceptable, then forced commutation must be used. The essence of forced
commutation is to decrease the SCR current to zero either by transferring the load current to a preferred path or by decreasing the load
current to zero.
5.3 CLASSIFICATION OF FORCED COMMUTATION METHODS
There are six distinct classes by which the energy is switched
across the SCR to be turned off:
Class A Self commutated by resonating the load
Class B Self commutated by an LC circuit
Class C C or LC switched by another load-carrying SCR
Class D C or LC switched by an auxiliary SCR
Class E An external pulse source for commutation
Class F AC line commutation
Examples of circuits which correspond to these classes will now
be given. These examples show the classes as choppers (Chapter 13).
The commutation classes may be used in practice in configurations
other than choppers. References to literature covering the different
classes will be found in Chapter 13.
5.3.1 Class A-Self commutated by resonating the load
When SCR1 is· triggered, anode current flows and charges up C
128
DYNAMiC CHARACTERISTICS OF SCR'S
in the polarity indicated. Current will then attempt to How through the
SCR in the reverse direction and the SCR will he turned off.
The condition for commutation is that the RLC circuit must he
under-damped.
FIGURE 5.9
CLASS A COMMUTATION
5.3.2 Class B-Self commutated by an LC circuit
Example 1
.
Before the gate pulse is applied, C charges up in the polarity
indicated.
When SCR1 is triggered, current flows in two directions.
1. The load current IR flows through R.
2. A pulse of current flows through the resonant LC circuit and
charges C up in the reverse polarity. The resonant-circuit current will
then reverse and attempt to flow through the SCR in opposition to the
load current. The SCR will tum off when the reverse resonant-circuit
current is greater than the load current.
191
C
SCR I
IR
L
E~
.=.
lIR
R
ISCRI
VSCRI
~
~
I ~
0,
}
I
C
I tCI
V
FIGURE 5.10
CLASS B COMMUTATION (EXAMPLE 1)
129
SCR·MANUAL
Class B-Self commutated by an lC circuit
Example 2 - The Morgan Circuit
From the previous cycle the capacitor is charged as shown in
Figure 5.11 and the reactor core has been saturated "positively."
When SCR1 is triggered, the capacitor voltage is applied to the
reactor winding L 2. The polarity of the applied voltage immediately
pulls the core out of saturation. For the time tl to t2 (Figure 5.11) the
load current is Howing through R. Simultaneously the capacitor is being
discharged.
When the voltage across L2 has been applied for t1;Ie prescribed
time, the core goes into "negative" saturation. The inductance of L2
changes from the high unsaturated value to the low saturated value.
The resonant charging of C now proceeds much more rapidly
from time t2 to ta. As soon as the peak of current is reached and current starts to decrease, the voltage across L2 reverses.
As soon as the voltage reverses, the core comes out of saturation
again, the inductance rises to the high value and the recharging of C
proceeds at a more leisurely pace (ta to t 4 ).
The voltage across the inductor is held for the prescribed time and
then positive saturation occurs (t4 ).
Now the capacitor is switched directly across the SCR via the
saturated inductance of L 2 • If the reverse current exceeds the load
current SCR 1 will be turned off. The remaining charge in C then is
I"
~
t\
to
ISCltl
v_,
+
SCR,
..
E~
_c
'----:§
+
POSITIVE
SATURATION
vc
CORE SATURATION
t=tF
I
!
...,t...
L
FIGURE 5.11
130
CLASS B COMMUTATION (EXAMPLE 2)
DYNAMIC CHARACTERISTICS OF SCR'S
dissipated in the load and C is charged up and ready for the next
cycle (t5)'
It is quite possible in practice to design L so that negative saturation does not occur. In this case the anode-current pulse from t2 to ta
is omitted.
5.3.3 Class C-C or LC switched by another load-carrying SCR
Assume SCR2 is conducting. C then charges up in the polarity
shown. When SCR 1 is triggered, C is switched across SCR2 via SCR1
and the discharge current of C opposes the flow of load current in
SCR2 •
R
R
+ -
seRa
Igi
1Rr
ISCRI
VSCR,
192
~
f\.
~
t
t
~
Itcl/
I
I
I-
b"
V
~
I
FIGURE 5.12
CLASS C COMMUTATION
131
SCR MANUAL
5.3.4 1:lass D-lC or Cswitched by an auxiliary SCR
Example 1
. The circuit shown in Figure 5.12 (Class C) can be converted to
Class D if the load current is carried by only one of the SCR's, the
other acting as an auxiliary tum-off SCR. The auxiliary SCR would
have a resistor in its anode lead of say ten times the load resistance.
Example 2
SCR2 must be triggered first in order to charge up the capacitor in
the polarity shown. As soon as C is charged, SCR2 will turn off due to
lack of current.
When SCR 1 is triggered the current Hows in two paths: Load current Hows in R; commutating current Hows through C, SCRlo L, and D,
and the charge on C is reversed and held with the hold-off diode D.
At any desired time SCR2 may be triggered which then places C across
SCR 1 via SCR2 and SCR1 is turned off.
~
19,
IR
1SCR,
VSCR,
+
c_
I
~C
I
SCR,
SCR2
I
1lIz
Ei
0
L
ISCRZ
R
VSCR2
Vo
Ie
FIGURE 5.13
132
.I
~
\
I
I:'
r
L
V
"'I
~l
I
D
.J
l/
~~
/'
Ie
L
./
V
CLASS 0 COMMUTATION (EXAMPLE 2)
I
"
...........
~
DYNAMIC CHARACTERISTICS OF SCR'S
Class D-LC or Cswitched by an auxiliary SCR
Example 3 - The Jones Circuit
The outstanding feature of this circuit is its ability to start commutating reliably.
If C were discharged, then, on triggering SCRlo voltage would be
induced into L z by closely coupled Ll> and C would become charged
in the polarity shown. As soon as SCRz is triggered, SCR1 turn-off interval is initiated. C now becomes charged in the opposite polarity.
The next time SCR 1 is triggered, C discharges via SCRb and L 2 ,
and its polarity is reversed ready for the next commutating pulse. The
voltage to which C is charged (in the polarity shown in Figure 5.14),
depends on which is greater: the voltage induced by load current
Howing in Ll or the reversal of the positive charge built up while SCR2
was conducting.
With heavy loads, the induced voltage increases, thus tending to
offset the decrease of turn-off time. Better turn-off times are obtained
with this circuit as compared with Example 2 at the cost of higher voltages appearing aCrOSS the SCR's. This circuit is discussed in more detail
in Chapter 13.
'"
I.
1 8CRI
SCRI~
,
9
VSCR1
~.
t:
~C
I
)C'
I
'"
I
L,
R
ISCR!
VSCR2
VLo
Vc
'.
FIGURE 5.14
r
I
~
~
I
~
f
,~
V
~
I
~~
I tc
V--
I
V
/
~
F V~~
CLASS D COMMUTATION (EXAMPlE 3)
133
8CH MANUAL
5.3.5 Class E-External pulse source for commutation
Example 1
When SCR 1 is triggered, current will How into the load. To turn
SCR1 off base drive is applied to the transistor Ql' This will connect
auxiliary supply E2 across SCR 1 turning it off. Ql is held on for the
duration of the turn-off time.
R
101
-Q
1 - - - 1
1.
VICRt
lBAK
11-------'-0-'------
IcoUEClOOIt-------,-,-h- -
J
5.15
134
CLASS E COMMUTATION (EXAMPLE 1)
DYNAMIC CHARACTERISTICS OF SCR'S
Class E-External pulse source for commutation
Example 2
The transformer is designed with sufficient iron and air gap so as
not to saturate. It is capable of carrying the load current with a small
voltage drop compared with the supply voltage.
When SCR 1 is triggered, current flows through the load and pulse
transformer. To tum SCR 1 off a positive pulse is applied to the cathode
of the SCR from an external pulse generator via the pulse transformer.
The capacitor C is only charged to about 1 volt and for the duration of
the turn-off pulse it can be considered to have zero impedance. Thus
the pulse from the transformer reverses the voltage across the SCR, and
it supplies the reverse recovery current and holds the voltage negative
for the required tum-off time.
I.,
lR
ISCRI
VSCRI
VPULSE
IpULSf
~
I
~
t
I
I
I
FIGURE 5.16
"L
l
D
I
f'-
0
!
CLASS E COMMUTATION (EXAMPLE 2)
135
SCR MANUAL
Class E-External pulse source for commutation
Example 3
When the SCR is turned on, the pulse transformer saturates and
presents a low impedance path for the load current. When the time
comes for turning off the SCR, the first step is to de-saturate the pulse
transformer. This is done by means of a pulse in the polarity shown.
This de-saturating pulse momentarily increases the voltage across the
load and also the load current. Once the pulse transformer is desaturated, a pulse in the reverse polarity is injected, reversing the voltage across the SCR and turning it off. The pulse is held for the required
tum-off time.
+
R
136
DYNAMIC CHARACTERISTICS OF SCR'S
Class E-External pulse source for commutation
Example 4
This circuit is important because no capacitor-charging pulse Hows
through the load.
Assume C is charged in the polarity shown to some voltage greater
than the supply voltage E. When SCR1 is triggered, load current Hows
in R and L 2 • SCR2 is in a resonant circuit consisting of C and L 2 • When
SCR2 is triggered, a pulse of current Hows through L 2 • A voltage is
developed across L2 which is greater than the supply voltage E. Reverse
voltage is therefore applied to SCR1 which turns it off. The termination
of the discharge pulse through SCR2 turns it off, and C is now charged
in the opposite polarity. Ll is much larger than L 2 , and C is now resonantly charged via Ll and D to some voltage greater than the supply
voltage.
L,
R
I
SCRI~
~~
~2
L2
Ig,
I.
I ....,
lisco,
tSCRz
VSeRZ
, +c
~
l
~
~r
~
(
l
r
1
V
I
~
It~
~a
v4r=
L7
I,
\
Yc
I
FIGURE 5.18
te
L7
~
CLASS E COMMUTATION (EXAMPLE 4)
137
SCR MANUAL
5.3.6 Class F-AC line commotated
If the supply is an alternating voltage, load current will How during the positive half cycle. During the negative half cycle the SCR will
tum off due to the negative polarity across the SCR. The duration of
the half cycle must be longer than the tum-off time of the SCR.
D
o,
/II;
LINE '"
R
10,
VSUPPLY
LO
j---'C-J
VeeRt
FIGURE 5.19
138
CLASS F COMMUTATION
DYNAMIC CHARACTERISTICS OF SCR'S
5.4 RATE OF RISE OF FORWARD VOLTAGE, dv/dt
The junctions of any semiconductor exhibit some unavoidable
capacitance. A changing voltage impressed on this junction capacitance
results in a current, i = C dv/dt. If this current is sufficiently large a
regenerative action may occur causing the SCR to switch to the onstate. This regenerative action is similar to that which occurs when gate
current is injected, as discussed in Chapter 1. The critical rate of rise
of off-state voltage is defined as the minimum value of rate of rise of
forward voltage which may cause switching from the off-state to the
on-state.
Since dv/dt turn-on is non-destructive, this phenomenon creates
no problem in applications in which occasional false turn-on does not
result in a harmful affect at the load. Heater application is one such
case.
The majority of inverter applications, however, would result in
circuit malfunction due to dv/dt turn-on. One solution to this problem is to reduce the dv/dt imposed by the circuit to a value less than
the critical dv/dt of the SCR being used. This is accomplished by the
use of a circuit similar to those in Figure 5.20 to suppress excessive
rate of rise of anode voltage. Z represents load impedance and circuit
impedance.
Since circuit impedances·are not usually well defined for a particular application, the values of Rand C are often determined by experimental optimization. A technique described in Chapter 16 can be used
to simplify snubber circuit design by the use of nomographs which
enable the circuit designer to select an optimized R-C snubber for a
particular set of circuit operating conditions.
Another solution to the dv/dt turn-on problem is to use an SCR
with higher dv/dt capability. This can be done by selecting an SCR
designed specifically for high dv I dt applications, as indicated by the
specification sheet; Emitter shorting, as discussed in Chapter 1, is a
manufacturing technique used to accomplish high dv/dt capability.
RS
(b) Circuit Variations
(a) Basic Circuit
FIGURE 5.20
DV/DT SUPPRESSION CIRCUITS
139
SCR MANUAL
Higher dv/dt capability can also be attained by choosing an SCR
with higher voltage classification. Since a high circuit-imposed dv/dt
effectively reduces V (BO) (the actual anode voltage at which the particular device being observed switches into the on state) under given temperature conditions, a higher. voltage classification unit will allow a
higher rate of rise of forward voltage for a given peak circuit voltage.
Alternatively, this increased dv/dt capability can be understood
by reference to Figure 5.21, the typical variation of dv/dt capability
with applied voltage. By choosing an SCR with higher VDR1\[, the ratio
of Vapplled to V DRlI will be lower for a given circuit and as indicated in
Figure 5.21 the typical dv/dt capability will be higher. By making use
of this technique, the SCR can be selected by the manufacturer for
dv/dt capability in excess of that indicated on the specification sheet.
IOIO~---2==O:---'---~40=--'--f.60::-.~80:!=-'-:-!IOO
VAPPLIEOI VORM -PERCENT
FIGURE 5.21
TYPICAL VARIATION OF DV/DT CAPABILITY
WITH APPLIED VOLTAGE
Reverse biasing of the gate with respect to the cathode may increase dv/dt capability for small area SCR's not already designed for·
high dv I dt. The reader is referred to Chapter 4 for further discussion.
5.4.1 Reapplied dv/dt
Reapplied rate of rise of voltage, reapplied dv I dt, is the rate of
rise of forward voltage following the commutation interval. Reapplied
dv I dt is of importance because of its affect on turn-off time. The affect
140
DYNAMIC CHARACTERISTICS OF SCR'S
on tq of reapplied dv I dt, a test condition for tq measurement, can be
seen in Figure 5.7. Triac applications, discussed in Chapter 7, also
require consideration of dv/dt.
5.5 RATE OF RISE OF ON-STATE CURRENT, di/dt
Critical di/dt is the maximum allowable value of the rate of rise
of on-state current. The di/ dt of on-state current while the SCR is in
the process of turning on must be considered because it is capable of
destroying the SCR or, in the absence of destruction, can cause a high
switching loss. During the turn-on process only a small percentage of
the silicon is conductive due to the finite spreading velocity, as discussed in Reference 2. A fast rising current can result in a high current
density in that portion of silicon that is conducting. This high current
density may result in excessive heat and a destroyed SCR.
5.5.1 Solutions to di/dt Problem
Inverter circuits with inherently high di/dt waveshapes can be
made to operate reliable by choosing an SCR with high di/dt capability. The SCR manufacturer can attain this high dil dt capability by
appropriate gate construction techniques as discussed in Chapter 1.
An additional technique used to accomplish high di/dt capability
is to employ a hard-drive gate circuit. A hard drive consists of a fast
rising gate current. Reference 5 discusses trigger circuit design techniques to accomplish high di/dt capability.
A saturable reactor in series with the SCR during its turn-on
switching interval will greatly reduce switching dissipation in the SCR.
When the SCR' is triggered on, the amount of current that will flow
during the turn-on interval is limited to the magnetizing current of the
reactor. The reactor is designed to go into magnetic saturation sometime after the SCR has been triggered. The delay time is employed to
bring operation of the SCR. within its tum-on current limit capability.
Sufficient SCR active area is then available to assume full load current
at minimum dissipation. Since the load current is delayed, the output
of the SCR-reactor combination is delayed relative to the SCR trigger
signal. Realistic pulse repetition rates are achievable by this technique,
notably in many pulse modulator applications.
The delay time t of the saturable reactor is given by the time to
saturate
NA 6B 10- 8
ts =
E
(seconds), where
(5.1)
N = number of turns
A = cross sectional area of core in square centimeters·
6B = total flux density change in Gauss
E = maximum circuit voltage being switched in volts
The current required at the time of saturable reactor switching I.
should be made small compared to the peak load current being
switched. It is:
141
SCR MANUAL
I•
=
H.lm
O.47TN
(amperes,) were
h
(3.4)
H. = magnetizing force'in Oersteds required for core flux
to reach saturation flux density B. (1 Oersted =
2.021 ampere-turns/inch)
1m = mean length of core in centimeters
N = number of turns
Provision must be made to properly reset the core before the next
current pulse, Depending on the details of the circuit, reset may be
accomplished by the resonant reversal of current (reverse recovery current) or by auxiliary means.
5;6 REVERSE RECOVERY CHARACTERISTICS
The time during which reverse recovery current flows in the SCR
(ts to t6 in Figure 5.1) is known as the reverse recovery time. This is
the time required before the SCR can block reverse voltage. This should
not be confused with tum-off time which is the time that has to elapse
before the SCR can block reapplied forward voltage. The reverse recovery phenomenon is also common in junction diodes.
Reverse recovery time in typical SCR's is of the order of a few
microseconds. Recovery time increases as forward current increases and
also increases as the rate of decay of forward current decreases. In addition, an increase of recovery time results from an increase of junction
temperature.
The reverse recovery current phenomenon plays a minor but important part in the application of SCR's:
1. In full wave rectifier circuits using SCR's as the rectifying elements the reverse recovery current has to be carried in the forward
direction by the complementary SCR's. This can give rise to high values
of turn-on current.
2. In certain inverter circuits such as the McMurray-Bedford circuit (Chapter 13) where one SCR is turned off by turning another on,
the reverse recovery current of the first gives rise to high values of
tum-on current in the second.
3. The cessation of reverse current, which can be very sudden,
may produce damaging voltage transients and radio frequency interference.
4. When SCR's are connected in series the reverse voltage distribution may be seriously affected by mismatch of reverse recovery
times (Chapter 6).
Recovered charge, QR, (the time integral of reverse current), is
the amount of charge in microcoulombs corresponding to the recovery
interval. Figure 5.22 indicates the dependence of recovered charge on
reverse di/ dt. An inductor may be needed to limit recovered charge
within a value acceptable for circuit operation. This inductance may be
in the form of source, circuit, or load impedance. This same inductor
serves to prevent di/dt failures as discussed in Section 5.5.1 above.
142
DYNAMIC CHARACTERISTICS OF SCR'S
10,00 0
I
II". I
C280/C'2.81
RECOVERED CHARGE
JUNCTION TEMP. 125-C
r--
500.11.,1000.11.
97% OF ALL DEVICES
IOOA~
BETWEEN MIN. AND MAle.
1--11
0
I-"'"
MAX•
./
1--'1-'
i-'"
./
0
M'~
00 •1
,
FIGURE 5.22
'0
REVERSE di/dt (AMPS/".SECl
'00
1000
RECOVERED CHARGE AT 125°C
5.7 CAPACITORS FOR COMMUTATION CIRCUITS
The characteristics of the commutation capacitors must be carefully considered by the design engineer in their selection and specification. The following properties are desirable:
1. The capacitor life should be long, at the operating ambient
temperature.
2. The power losses in the capacitors should be low for two
reasons:
a. To avoid high internal temperatures which would shorten
the capacitor life.
b. To maintain the advantage of high efficiency which the SCR
gives to the over-all circuit.
3. The capacitor's equivalent series inductance should be known.
In many circuits, inductance in series with the commutating
capacitor plays an important part in controlling the initial rate
of rise of anode current through the SCR.
The equipment designer is advised to take the following steps:
1. For the breadboard, standard inverter capacitors may be purchased from the General Electric Capacitor Department, Hudson Falls,
New York. Figure 5.24 gives the ratings of standard capacitors. All
General Electric SCR capacitors have heavy duty internal connections
to carry the high currents, extended foil construction to give low inductance and minimum ESR (equivalent series resistance) and incorporate painted cases to keep the dielectric temperature rise to a
minimum.
2. After completion of the breadboard tests, the voltage and current waveforms and temperature data should be submitted to the
capacitor manufacturer for optimization of life, size, and cost. (See
check list at end of chapter.)
The RMS current encountered in SCR applications is usually significant and even with minimum ESR the J2R losses can be great. While
143
SCR MANUAL
proper capacitor selection will provide a suitable component, the inherent power losses must be considered by the designer from a total circuit
standpoint.
Another important consideration is the current carrying capability
or limit of the capacitor itself. The maximum current capability of any
capacitor listed in the Standard Ratings Table is 50 amperes RMS.
Metalized designs have limits of 12 and 20 amperes RMS to gain optimum volumetric efficiency. In several cases, the RMS current listed is
greater than these actual operating current limits. The greater values
may only be used for derating purposes along with multipliers from the
derating table Figure 5.23. All capacitors with such listed ratings are
clearly indicated in the Standard Ratings Table with explanatory
footnotes.
a
5
I
~
r--.... r---t--
IRMS=IPKA
.......
5
'" . . . r---..
.110
50
100
500
r-.r-.
1000
5000
10.000
I," CAPACITOR CHARGE ANDIOR DISCHARGE TIME (MICROSECONDS)
FIGURE 5.23
CORRECTION FACTOR TO BE APPLlEO TO CURRENT (IRMs) FOR CAPACITOR CHARGE
OR DISCHARGE TIMES OTHER THAN 50 MICROSECONDS
The RMS current values in the standard ratings table are based
on a current pulse width of 50 microseconds. Selection of capacitors for
circuits involving 50 microseconds pulse widths can be made directly
from the Standard Ratings Table by knowing the capacitance, voltage,
RMS current and ambient operating temperature. For circuits involving
pulse widths other than 50 microseconds the following example is
offered.
Pulse width less than 50 p:;ec. Given:
Capacitance: 5 MFD ± 10 percent
Voltage: 65 VAC, 16.6 Khz continuous sinewave
Current: 34 amps RMS
Temperature: 80 C
a. Select a 5 MFD unit with an 80 C current rating near 34 amps
RMS. 28F1248 has a current rating of 30.6 for 50 microseconds.
b. The current multiplier for 16.6 Khz (pulse width = 30 microseconds) from Figure 5.23 is 1.2.
c. Allowable current for Cat. No. 28F1248 at 16.6 Khz is (1.2)
(30.6) = 36.7 amps RMS.
d. Since the allowable current of 36.7 amps is greater than the
requIred value of 34 amps RMS, this unit is adequate.
144
DYNAMIC CHARACTERISTICS OF SCR'S
STANDARD RATINGS
Nameplate Rating
Dielectric
Paper
Paper
Paper
Paper
Paper
Paper
Polycarbonate
Metalized
Paper
Dimensions in inches
Volts
Max.
Case RMS
Depth Height AC
Peak
Volts·
JLf
Catalog
Number
200
200
200
200
200
200
200
200
200
200
1
2
3
5
10
15
20
30
40
50
28F5101
28F5102
28F5103
28F5104
28F5105
28F5106
28F5107
28F5108
400
400
400
400
400
400
400
400
400
400
1
2
3
5
15
20
30
40
50
28F5110
28F5111
28F5112
28F5113
28F5114
28F5115
28F5116
28F5117
28F5118
600
600
600
600
600
600
600
600
600
600
600
1
2
3
5
10
15
20
25
30
40
50
28F5120
28F5121
28F5122
28F5123
28F5124
28F5125
28F5126
28F5127
28F5128
28F5129
2.16
2.16
2.16
2.91
2.91
3.66
3.66
3.66
4.56
4.56
1.31
1.31
1.31
1.91
1.91
1.97
1.97
1.97
2.84
2.84
1000
1000
1000
1000
1000
1000
1
2
3
5
10
20
28F5131
28F5132
28F5133
28F5134
28F5135
28F5137
2.16
2.16
2.16
2.69
2.91
4.56
1500
1500
1500
1500
1500
1500
0.5
1
2
3
5
10
28F5141
28F5142
28F5143
28F5144
28F5145
28F5146
2000
2000
2000
2000
2000
2000
2000
0.25
0.50
1
2
3
5
10
600
600
600
600
600
600
200
200
200
200
200
Width
Max. RMS (Amperes)
At Max. Ambient Tempt
60 C
70 C
200
200
200
200
200
200
200
200
200
200
8.9
12.8
22.1
30.1
37.7
52.6:1:
68.9:1:
81.1:1:
6.6
9.5
16.3
22.1
27.8
38.9
50.9:1:
59.9:1:
3.8
5.5
9.4
12.8
16.1
22.4
29.4
34.6
Use 1000 Volt Rating
300
300
300
300
300
300
300
300
300
300
7.3
9.4
14.0
26.6
35.4
46.9
65.3:1:
82.1:1:
89.4:1:
5.4
7.0
10.3
19.7
26.2
34.6
48.2
60.6:1:
66.0:1:
3.1
4.0
6.0
11.4
15.1
20.0
27.8
35.0
38.1
Use 1000 Volt Rating
2.31
2.81
3.81
4.25
5.75
5.75
6.75
8.00
5.88
6.75
400
400
400
400
400
400
400
400
400
400
400
7.7
10.4
15.6
27.3
39.0
49.2
59.6:1:
71.1:1:
79.1:1:
95.8t
5.7
7.7
11.5
20.2
28.8
36.3
44.0
52.5:1:
58.9:1:
70.8:1:
3.3
4.4
6.7
11.7
16.6
21.0
25.4
30.3
33.7
40.8
1.31
1.31
1.31
1.56
1.91
2.84
2.06
3.06
3.81
4.25
6.25
5.18
500
500
500
500
500
500
5.1
8.9
12.1
18.3
33.2
52.8:1:
3.8
6.5
8.9
13.5
24.5
39.0
2.2
3.8
5.2
7.8
14.1
22.5
2.16
2.16
2.69
2.91
2.91
4.56
1.31
1.31
1.56
1.91
1.91
2.84
2.06
3.06
3.88
4.25
6.25
5.18
700
700
700
700
700
700
3.6
6.3
11.0
15.0
23.5
37.3
2.7
4.6
8.2
11.1
17.3
27.6
1.5
2.7
4.7
6.4
10.0
15.9
28F5151
28F5152
28F5163
28F5154
28F5155
28F5156
28F5158
2.16
2.16
2.16
2.69
2.91
3.66
4.56
1.31
1.31
1.31
1.56
1.91
1.97
2.84
2.06
2.56
3.44
4.50
4.75
6.25
6.25
800
800
800
800
800
800
800
2.6
4.0
6.6
11.9
15.8
25.6
41.2
1.9
3.0
4.9
8.8
11.7
18.9
30.4
1.1
1.7
2.8
5.1
6.8
10.9
17.6
1
2
3
5
10
20
28F1245
28F1246
28F1247
28F1248
28F1249
28F1202
2.16
2.16
2.16
2.69
2.91
3.66
1.31
1.31
1.31
1.56
1.91
1.97
2.06
2.81
3.44
3.88
5.25
6.00
330
330
330
330
330
330
21.0
34.0
45.6
72.0:1:
120.0:1:
189.0:1:
15.5
25.3
33.8
53.0:1:
89.5:1:
148M
9.0
14.6
19.5
30.6
52.0:1:
86.0t
25
50
100
125
150
28Fll01
28Fll02
28Fll03
28Fll04
28F1105
2.69
2.69
3.66
3.66
3.66
1.56
1.56
1.97
1.97
1.97
2.12
2.88
3.12
3.88
4.25
120
120
120
120
120
12.5§
20.4§
34.61)
43.21)
49.51)
9.7
15.8§
26.81)
33.21)
38.21)
4.7
7.7
13.0
16.3
22.01)
10
1~00
Use
VOltlRating
Use 400 Volt Rating
FIGURE 5.24
2.16
2.16
2.16
2.16
2.69
2.91
2.91
3.66
2.16
2.16
2.16
2.69
2.91
2.91
3.66
3.66
4.56
1.31
1.31
1.31
1.31
1.56
1.91
1.91
1.97
1.31
1.31
1.31
1.56
1.91
1.91
1.97
1.97
2.84
2.06
2.56
3.81
4.69
4.50
5.25
6.75
6.25
2.06
2.31
3.06
4.50
4.75
6.25
6.75
8.00
5.88
80 C
EXTRACTS FROM GE CAPACITOR CATALOG
145
SCR MANUAL
* See
"Max. RMS Volts AC" column for a·c Rating.
Based 00 50 microsecond'pulse width.
:j: This number is given..for purposes of derating only, In no case may capacitor be operated at
currents in excess of 50 amps RMS.
§ This number is given for purposes of derating only. In no case may capacitor be operated at
currents in excess of 12 amps RMS.
11 This number is given for purposes of derating only. In no case may capacitor be, operated, at
currents in excess of 20 amps RMS.
t
CAPACITORS FOR SCR COMMUTATION APPLICATIONS
Design Data Sheet
To assist in obtaining proper capacitor design, it is particularly
important that the circuit design engineer sketch out in detail a picture
of voltage and current vs. time. This should be done by using the space
provided below and showing specific values of voltage, current, and
time over a complete cycle.
Primary Information:
Reference No. ______
Tolerance (if less than
± 10 percent,-_ _ _ __
Peak to Peak Voltage:
Rms Voltage: _ _ _ _ __
Peak Current:
Rms Current: _ _ _ _ __
Repetition Rate:
Duty Cycle: _ _ _ _ _ __
(time on - time off)
(cycles per second)
Ambient Temperature,___' _ _ _Max. ____Min. _ _ __
Capacitor Discharge Time: _ _ _ _ _ _ _ _ _ _ _ _ __
Show Sketch of Voltage and Current Wave shapevs. Time.
(Fill in below)
1. Capacitance Required: _ __
2.
3.
4.
5.
6.
7.
+
Volts 0
Time
+
Current 0
146
Time
DYNAMIC CHARACTERISTICS OF SCR'S
Secondary Information:
8. Desired Operating Life: _ _ _ _ _ _ _ _ _ _ ,(total cycles)
_ _ _ _ _ _ _ _ _ _,(total hours)
9. Sample Requirements: (How many)_ _ _ _ _ _ _ _ _ __
(When needed) _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __
10. Potential Usage: _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __
11. Physical Size Limitations: _ _ _ _ _ _ _ _ _ _ _ _ __
12. Mounting Requirements: _ _ _ _ _ _ _ _ _ _ _ _ __
13. Applicable Specifications (if any): _ _ _ _ _ _ _ _ _ __
14. Unusual Atmospheric Conditions: _ _ _ _ _ _ _ _ _ __
(dust, chemicals, humidity, corrosion, etc.)
15. Other Special Requirement: _ _ _ _ _ _ _ _ _ _ _ __
(high altitude, shock, vibration, etc.)
16. What kind cooling available: _ _ _ _ _ _ _ _ _ _ _ __
(fins, heat sink, forced air, etc.)
REFERENCES
l. D. E. Piccone and 1. S. Somos, "Are You Confused by High di/dt
SCR Rating?", The Electronic Engineer, January 1969, Vol. 28,
No. l.
2. Application Note 200.28, "The Rating of SCR's When Switching
Into High Currents," N. Mapham, May 1963.
3. S. J. WU, "Analysis and Design of Optimized Snubber Circuits for
dv/dt Protection in Power Thyristor Applications" presented at
IEEE IGA Conference, October 1970. Available from General
Electric Publication 660.24.
4. GE Capacitor Catalog GEA-8688.
5. J. M. Reschovsky, "Design of Trigger Circuits for Power SCR's,"
GE Application Note 200.54, February 1970.
147
SCR MANUAL
NOTES
148
SERIES AND PARALLEL OPERATION
6
SERIES AND PARALLEL OPERATION
Since the introduction of SCR's in 1957, the power handling capability has been steadily improving with enhanced dynamic performance.
SCR's with rated blocking voltage to 2600 volts and RMS current to
llOO amperes are readily available today. The power handling capability of an SCR appears to be limited by the effective utilization of
larger and larger silicon wafers, methods of packaging, and techniques
of cooling the junction temperature. Still, there are numerous applications where a single SCR cannot meet the power requirements, such
as in terminals for HVDC transmission lines and rapid transit systems
where system requirements dictate operation at higher voltage and/or
current than can be realized within the capabilities of a single SCR.
Series/parallel combinations must be employed if system requirements
are to be met.
When SCR's are connected in series for high voltage operation,
both steady-state and dynamic-state voltages must be equally shared
across each unit. The di/dt and the dv/dt limitations must be assured
not to exceed the ratings of each SCR. When SCR's are connected in
parallel to obtain higher current output, the equalization of forward
current, both during the tum-on interval and the conduction state,
must be guaranteed either by matching the forward characteristics of
individual units or by employing external forced sharing techniques.
Little work has been done on series/parallel connection of triacs.
It appears the guidelines outlined here for SCR's generally apply to
triacs. To date, the triac has been primarily employed in consumer/
appliance/light industrial applications where slightly lower cost of
trigger and control circuitry can be important factors.
As requirements for operation from higher voltage sources and
at higher current levels, the applications may very well fall in the
realm of heavy industrial applications. Therefore, inverse parallel SCR's
are in general more appropriate to provide the function of AC switching.
6.1 SERIES OPERATION OF SCR's
When circuit requirements dictate operation at a voltage in excess
of the blocking voltage capabilities of a single SCR, series combinations
can be employed if certain design precautions are taken. These precautions are primarily the equalization of voltage sharing, both forward
and reverse, between individual SCR's at steady-state and transient
operation conditions. Due to differences in blocking currents, junction
capacitances, delay times, forward voltage drops as well as reverse
recovery for individual SCR's, external voltage equalization networks
and special consideration in gating circuits design are required.
149
SCR MANUAL
6.1.1
Need For Equalizing Network
Shown in Figure 6.1 are
hypothetical voltage / current
characteristics for two randomly selected SCR's. If the
two SCR's are connected in
series one might expect them
to have a total forward blocking capability of at least 2 (V2)'
Yet without forced voltage
equalization the total peak
forward blocking voltage must
be limited to approximately
(VI + V2) in order to keep the
voltage across SCR2 from exceeding V WO )2'
I(BOJ,
-
SCR ANODE VOLTAGE _
FIGURE G.1
SCR CHARACTERISTICS
Figure 6.2 shows, diagramaticaIly, the six operating states that
can occure in a random sample of SCR's, connected in series, without
forced equalization. It is seen that the equivalent individual impedances
change continuously as the SCR series connection switches from state
to state.
:nr
m
II
FORWARD
BLOCKING
PARTIAL
-+
TURN -ON
FORWARD
---Joo
:m:
JZ:
REVERSE
OONDUCTING
CONDUCTING
---+
PARTIAL
REVERSE
REV RECOVERY
BLOCKING
+1200 VOL.TS
SCR,
1000V
1200V
I.OV
O.9V
0.7'11'
100V
50V
.• V
I.IV
1.0'11'
0.7V
900V
I50V
OV
O.BV
12.00'11'
200V
seR2
0.9'11'
SCR3
~
5MA
FIGURE 6.2
~
IQMA
1- 1
50A
lOA
-
t
IOMA
t
IOMA
POSSIBLE OPERATING STATES OF AN UNEQUALIZED SERIES STRING OF SCI'S
During forward and reverse blocking states (I and VI) the difference in blocking characteristics result in unequal steady-state voltage
sharing. This could be harmful to an SCR with inherently low blocking
current since it might cause excessive voltage to appear across that SCR
under blocking states. In order to equalize the voltage, a shunt resistor
is connected across each SCR.
150
SERIES AND PARALLEL OPERATION
The conduction states (III and IV) represent no problem of voltage equalization.
States II and V represent undesirable unbalanced transient voltage
sharing during turn-on and reverse recovery conditions.
In state II, the delay time of one SCR is considerably longer than
other SCR's in the series string, consequently full voltage will be
momentarily supported by the slow turn-on SCR. One method that
can be taken to minimize unbalance caused by dissimilar turn-on delays
is to supply high enough gate drive with fast rise time to minimize
delay time differences. State V results from the fact that in a randomly chosen series string of SCR's, all SCR's will not recover at the
time instant. The first cell to regain its blocking voltage capability will
support the full voltage. To equalize the voltage during this period,
a capacitor is connected across each SCR. If the impedance of the
capacitor is low enough, and/or the time constant is properly chosen,
the voltage buildup on the fastest SCR to recover is limited until the
slowest one also recovers. This also alleviates the undesirable condition
of state II.
In summary, states III and IV present no equalization problem.
Shunt resistors equalize the voltage during states I and VI. Shunt
capacitors equalize the voltage during states II and V. High gate-drive
reduces inequalities during state II.
While capacitors provide excellent transient voltage equalization,_
they also produce high switching currents through the SCR's during
the turn-on intervaJ.1,2 Switching currents can be limited by means of
damping resistors in series with each capacitor. Although it is desirable
to have a large value of R and therefore a small value of C to limit the
power dissipation in the RC circuit, the value of the damping resistors
must be kept to a reasonably low value in order not to reduce the
effectiveness of the capacitors in equalizing voltage during the reverse
recovery interval. Also, low values of damping resistance preclude
excessive voltage build-up due to the IR drop during How of reverse
recovery current in the series connection after the first SCR has
recovered.
Figure 6.3 shows the voltage equalization scheme described above.
I STATIC
I EQUALIZING
NETWORK
I
"
"
FIGURE 6.3
SERIES EQUALIZING ARRANGEMENT
151
SCR MANUAL
Diodes can be placed across the damping resistors Rn to increase the
effectiveness of the capacitors in preventing misfiring due to excessive
rate of rise of forward voltage on the SCR·s. Some small amount' of
damping resistance should still be used in series with the diode to
prevent ringing.
It is cautioned that diodes should have soft recovery characteristics; otherwise, the abrupt recovery action of a snappy diode may
produce an adverse effect, such as high voltage spikes and therefore
hinder the performance of the RC circuitS. 3 ,4
6.1.2 Equalizing Network Design
6.1.2.1 Static Equalizing Network
For any given random group of SCR's there will be a given range
of forward and reverse blocking current at given circuit conditions.
Naturally, SCR's with low inherent blocking current will assume a
greater portion of a steady state blocking voltage than will units with
higher blocking current when all are connected in series. If the range
of blocking current is defined as Ib(max) - Ib(min)
a Ib' it is seen
that the maximum unbalance in blocking voltage to SCR's of a series
string occurs when one member has a blocking current of Ib(min) and
all remaining SCR's have Ib(max). Figure 6.4 represents just such a case.
=
- ... +
-- ------ -- ----...;::1=:- -- -- --- --- - - - - Eb
;---------+.,'1'1.=----------\
4-+
Ib(min)
IblmG.J
R,
R.
......
......
I- "
E.
I'
FIGURE 6.4
152
---01
'2
Ib (max)
Ib(max)
R.
R,
4-+
4-+
'2
Em
'2
'1
USE OF SHUNT RESISTORS TO EQUALIZE BLOCKING VOLTAGES TO
SERIES SCR'S
SERIES AND PARALLEL OPERATION
Choose Ep as the maximum blocking voltage which we will allow
across anyone SCR. By inspection 11 > 12 , Therefore:
Ep = 11 R.
Also:
where:
Em
n.
12
Em
= peak blocking voltage to entire series string
= number of SCR's in series. string
= 11 - ~Ib
= Ep + (ns - 1) R. (11 - ~Ib)
= n.Ep - (n. - 1) R. ~Ib
Now:
n.Ep - Em
R.:2i (n. _ 1) ~Ib
(6.2)
In general, only the maximum blocking currents for a particular
SCR type are provided by the manufacturer. If one wishes to be conservative, Ib(min) can be assumed to be zero. The required value for Rs
then becomes
(6.3)
Equalization resistors represent power consumers and as such it
is desirable to use as large a resistance as possible. In :3 given group
of SCR's, chances are' good that one can select ~Ib to be considerably
less than Ib(max). For this reason the alb approach is recommended.
When determining ~Ib' it is best to measure blocking currents at maximum rated junction temperature and blocking voltage. After ~Ib groups
are selected, the ~Ib should be checked at 25°C. To allow for differences in SCR temperatures when operating; a safety factor on ~Ib
should be used for design purposes.
Up to this point nothing has been said whether one must consider
forward or reverse blocking current, or both. In general an SCR specification sheet gives one figure to cover both. forward and revers.e blocking current; when both are specified, they are usually the same.
Figure 6.5 is a useful aid for finding therequir.ed voltage equalizing resistance for series strings up to eight SCR's long. Enter the
chart with a known Em/Ep and read up. to the curve designating
the number of SCR's per string. Read across to find E R. I . With a
knowledge of Em and ~Ib' determine R.(max).
m/~
b
153
·SCR MANUAL
0.5~~--+-----+-----~----~-----+-----;----~
R s'
-
MAX IMUM SHUNT RESISTANCE
Em - PEAK VOLTAGE TO SERIES STRING.
0.4
t-+--'t-tr--
E. _ PEAK VOLTAGE TO ANY SCR
.n. - NO. OF SERIES
SCR'S
4 Ib - RANGE OF BLOCKING CURRENT
r.
0.2
1~
0.1
Em
_
Ep
FIGURE 6.5
VOLTAGE EQUALIZING RESISTANCE FOR SERIES OPERATION OF SCR'S
To determine the power rating of the shunt resistors one must
consider the resistor which experiences the highest peak voltage. The
resistor effective power dissipation can be expressed as:
(EltMS)2
(6.4)
Rs
154
SERIES AND PARALLEL OPERATION
For phase control applications the maximum power dissipation occurs
at zero conduction angle:
~
(6.5)
For square wave applications:
JlJL
~t~
P
D
=~(_t)
Rs
T
(6.6)
For triangle wave applications:
~
~tl+-t,? ::::!-I
(6.7)
6.1.2.2 Dynamic Equalizing Network
As mentioned earlier, shunt capacitors are required to limit rate of
rise of voltage on the SCR's. Also during the reverse recovery interval
such capacitors provide a reverse recovery current path for slow SCR's
around those SCR's which recover first. Since the problem we are
trying to correct arises from a difference .in recovery characteristics
within a given type of SCR we must take some time to discuss this
characteristic. 3 ,4,5
Two SCR's with a sizeable difference in reverse recovery current
are depicted in Figure 6.6. The difference in the enclosed area is the
differential charge (designated .6.Q).
"TIME ZERO"
~
i
TIME_
SCR,
I
RM2
--------
FIGURE 6.6
REVERSE RECOVERY CURRENTS FOR TWO UNMATCHED SCR'S OF
THE SAME TYPE
155
SCR MANUAL
Note that Figure 6.6 shows ts2 > t rrl . This will not necessarily be
true for two randomly chosen SCR's. However if one has two SCR's
which represent the limit cases for a given type (i.e. worst case reverse
recovery mismatch) then t.2 is generally greater than trrl' For design
purposes this is a valid assumption.
Figure 6.7 shows current and voltage waveforms, during the
reverse recovery interval, for two· mismatched, series conn~cted SCR's
with shunt capacitors.
FIGURE 6.7
RECOVERY VOLTAGES OF CAPACITORS SHUNTING MISMATCHED,
SERIES CONNECTED SCR'S
From to to tl both SCR's present a short circuit to the How of
reverse current. Reverse current is applied at a rate determined by the
commutating voltage Ee and the circuit commutating reactance. Le·
During the period tl to t 2, capacitor C 1 begins to charge as SCRl begins
to regain its blocking capability. The rate of charge of C 1 increases
during this interval as the current in SCR 1 "tails-off." From t2 to ts
SCR 1 has fully recovered. The voltage and rate of charging C 1 further
increases due to the increasing reverse current in SCR2 • At t s, SCR2
begins to recover and the current through SCR2 begins to decrease
thus reducing the charging rate of Cl' C2 begins to charge as soon as
SCR2 begins to regain its blocking ability. At time t4 both SCR's are
fully recovered so that the only current path is through the capacitors.
This means that at time t4 the slopes of the voltage waveforms must
be equal. From time t4 on, the circuit behaves as a simple LRC circuit
(C being a series combination of n discrete capacitors). Due to the
difference in SCR recovery characteristics, the shunt capacitors when
charged to peak voltage are not charged equally. The maximum difference in voltage is designated as ~VmaX' This ~Vmax can be simply
represented by
156
SERIES AND PARALLEL OPERATION
V
.:l max
= .:lQmax
C
= 1,2,3 .... n
(B.8)
x
X
For steady state reverse blocking the shunt resistors share voltage
to the designed degree. In Figure B.7, the voltage difference between
the two shunt capacitors varies from .:lVmax at t5 to that determined by
the shunt resistors at some time later than t5' Assuming that the resistors
share the steady state reverse voltage perfectly, .:lV after t5 can be
represented by:
.:lV
.:l V max € - [t/R.C]
(B.9)
(time zero at t 5)
The worst combination of recovery characteristics for a series
string of SCR's is with one fast recovery SCR and all the remaining
ones being the slowest of that type. In this situation the peak voltages
across the shunt capacitors are as follows:
(B.10)
V0 (fast SCR)
(lin.) [Eo + (n. - 1) .:lVmax]
V0 (slow SCR)
(lin.) (Eo - .:lVmax)
Using relationships (B.8) and (B.10) and setting V0 (fast SCR)
Ep
=
=
=
c:::::
=
1) .:lQmax
(B.ll)
nsEp - Eo
One might say: "This relationship for C is fine but how do I find
.:lQmax for a given type of SCR"? Following is a table of typical spread
of .:lQ for some General Electric SCR types.
(n s -
SCH
C35
C137
C139
C140
C150
C154
C158
C180
C185
C280
C290
C398
C50l
Related Types
AQmax I"Coulombs
C3B, C37, C38, C40
C13B
C144
C141
C350
C155, C354, C355
C358
C380
C385
C281,C282,C283,C284
C291, CBOO
C387, C388, C397
CB01
CB02
TABLE 6.1
1
2
1
.5
30
18
22
45
20
400
340
40
400
400
TYPICAL RECOVERY CHARGE DIFFERENCE
TJ
= 125°C di/dt = 10A//LS
The reverse recovery charge is a function of both the device design
characteristics and the circuit commutation conditions. By varying the
commutation conditions, such as the magnitude of forward conduction
current, circuit inductance and the device junction temperature, the
recovery charge will vary accordingly. The values for .:lQ shown in
Table B.1 are for rated forward current at rated maximum junction
temperature and a rate of reverse current of 10 AI/Lsecond. For detailed
information other than specified, the reader is advised to refer to the
individual SCR data shc;ets. The corresponding data sheet numbers are
listed in Chapter 22.
157
SCR MANUAL
Note that up to this point we have talked about voltages across
the capacitors and tacitly assumed that such voltages are those on the
SCR's. In practice this is usually not true. As shown in Figure 6.8 a
certain amount of stray inductance is found in the physical capacitorSCR loop.
Le
+
(a)
FIGURE &.8
(b)
STRAY INDUCTANCE OF SCR·CAPACITOR LOOP
During the period t f (Figure 6.8), current is changing in such a fashion
as to set up a voltage in the stray inductance as shown. Realizing that
-this inductance is composed of wiring inductance, capacitor inductance
and that inherent in the SCR, it is not hard to visualize, say, 1 ph in
the loop. The induced voltage as shown represents additional reverse
voltage to the SCR. During t f the current can be changing at the rate
of 200 amperes per microsecond when the commutation conditions are
severe. This (with 1 ph) would mean an additional 200 volts reverse
to the SCR. It cannot be overemphasized that the inductance of the
capacitor-SCR loop should be held to as low a value as possible. For
this reason GE extended foil capacitors are suggested.
Since the shunt capacitors discharge through the SCR's during
turn-on, it is necessary to insert a small amount of resistance in series
with each capacitor. The value of the resistance is chosen to limit the
discharge current within the turn-on current limit of the SCR. Usually,
the required value of resistance will fall between 5 and 50 ohms. In
addition to limiting capacitor discharge current, high frequency oscillations due to interaction between the capacitors and circuit inductance
are suppressed. It must be remembered that, although the damping
resistance must be large enough to limit turn-on current and dildt, it
must not be so large that it either destroys the effect of the shunt
capacitors or establishes an excessively high voltage during flow of
reverse recovery current through it.
The value of this resistance can be estimated by the following
formula:
RD
K C/L
(6.12)
where
K is a function of 'allowed overvoltage and circuit parameters 3 typical value is in the range of 1.25 to 1.5.
=
158
SERIES AND PARALLEL OPERATION
Methods of designing and selecting the proper damping resistor
are discussed in Chapter 16.
&.1.2.3 Other Voltage Equalizing Arrangements
The arrangement of Figure 6.3 provides voltage sharing under all
conditions of forward and reverse blocking. In applications where the
increase in blocking losses due to current through the equalizing
resistors must be avoided, as in SCR radar modulator switches, voltage
sharing may be successfully accomplished by replacing each shunt
equalizing network with a silicon controlled avalanche rectifier as
shown in Figure 6.9(a). When maximum avalanche voltage is chosen
correctly, total forward blocking current through the circuit need be
only slightly higher than the maximum blocking current of the worst
SCR. Maximum avalanche voltage of the shunt rectifier should be
equal to, or slightly below, the SCR forward breakover voltage specifiG.E. CONTROLLED AVALANCHE RECTIFIERS
I :1 ;: I
VOLTAGE SHARING UNDER FORWARD BLOCKING. NO
REVERSE
BLOCKING CAPABILITY
(8)
VC~TAGE
SHARING UNDER BOTH FORWARD
REVERSE
BLOCKING
AND
VOLTAGE SHARING UNDER FORWARD BLOCKING.
REVERSE BLOCKING BUT NO PROVISION FOR
REVERSE SHARING
(b)
(e)
GE-MOV1IIC
BIDIRECTIONAL VOLTAGE SHARING
WITH GE-MOV
(d)
FIGURE 6.9
SERIES EQUALIZING USING CONTROLLED AVALANCHE
RECTIFIER AND METAL·OXIDE VARISTORS
159
SCR MANUAL
cation. Minimum avalanche voltage must be higher than Emlns when
measured at the controlled avalanche rectifier's minimum operating
temperature. To provide optimum equalization it is desirable to have
as narrow a tolerance as possible on the avalanche voltage of the shunt
rectifier. Where a series string has to block appreciable reverse as well
as forward voltage, inverse series controlled avalanche rectifiers may
be substituted for the single units (see Figure 6.9(b». In cases where
reverse blocking requirements are not severe, some reverse blocking
ability can be obtained using controlled avalanche rectifiers and conventional silicon rectifiers arranged as in Figure 6.9(c).
F.j$.ure 6.9(d) shows the shunt equalizing network utilizing a CEMOV which provides a sharp voltage clip in both forward and reverse
directions. The MOV'l"Mfunctionally replaces two series connected
avalanche diodes. Refer to Chapter 16 for more information regarding
the GE-MOV:""
6.1.3 Triggering Series Operated SCR's
There are two primary methods in common use for triggering
series SCR's, namely:
1. Simultaneous triggering
2. Slave triggering whereby one "master" SCR is triggered, and
as its forward blocking voltage begins to collapse, a gate signal
is thereby applied to the "slave" SCR.
Simultaneous triggering of all SCR gates is the preferred method.
Slave triggering, while it is a unique way to provide gate isolation,
produces some time delay between master and slave. Fortunately the
capacitors used for voltage equalization during the reverse recovery
period also limit the forward voltage rise. As long as the shunt capacitance is sufficient to limit forward voltage within the PFV ratings of
the SCR's until all SCR's are "on," slave triggering can be reliably
employed. The designer is cautioned to observe gate drive requirements of the SCR when employing slave triggering, particularly if
switching into a fast rising anode current.
6.1.3.1 Simultaneous Triggering Via Pulse Transformer
When using pulse transformers particular attention should be
given to the insulation between windings. This insulation must be able
to support at least the peak of the supply voltage.
Triggering requirements may differ quite widely betw~enindi
vidual SCR's. To prevent a cell with a low impedance gate characteristic from shunting the trigger signal away from a cell with a high
impedance gate characteristic, resistance should be inserted in each
gate lead, or impedance built into the transformer via leakage reactance,
as shown by Rg in Figure 6.10.
Where the total energy available to trigger is limited, as may well
be the case in a pulse triggering arrangement, it is preferable to replace
these resistors with capacitors in series with each gate lead. Series
capacitors tend to equalize the charge coupled to each SCR gate during
trigger pulses, thus reducing the effects of unequal loading without
160
SERIES AND PARALLEL OPERATION
"s
",
C,
seRI
(+)
C2
"2
SCA 2
~
.~
I-l
~
'0(
"gk
.,
",k
.ftC,
--,
----,
,
*,
I
g
C
"g
Cg
=*
I
___ J
I
- -"
'-
n
FIGURE 6.10
SIMULTANEOUS TRIGGERING OF SERIES CONNECTED SCR'S VIA
PULSE TRANSFORMERS
additional energy dissipation. When capacitors are used in this manner
a resistor, Rgk , should be connected from gate to cathode of each SCR
to provide a discharge path for the capacitor. The circuit must be able
to pass a fast rise time pulse, preferably less than 1 ,usecond.
It must be emphasized again that marginal triggering is discouraged. Most SCR specification sheets today show a preferred triggering
area on the gate characteristics curve. Particularly when switching into
high currents, operation below this preferred area can be disastrous.
6.1.3.2 Simultaneous Triggering by Means of Light
Figure 6.11 shows an approach whereby simultaneous triggering
of series connected SCR's is achieved by triggering LASCR's in the gate
circuit of each SCR. This method of triggering provides the required
gate isolation along with simultaneous tum-on when a single light
source is used to tum on all LASCR's. The series combination of RI
and R2 is made. equal to the required shunt resistance Rs. R2 is made
fairly small compared to RI so that low voltage LASCR's can be
employed. The RIC I time constant must be made sufficiently small so
that C I is fully charged to the voltage dictated by R2 at tum-on. Resistor
R4 limits the peak gate current. A useful trigger circuit (for LASCR's)
employing Xenon flashtubes is shown in Figure 6.12. The circuit as
shown operates well in the 60-400 Hz frequency range. The unijunction transistor relaxation oscillator provides alternate trigger pulses to
the two C5 SCR's. As the C5 SCR's tum-on, the .22 pi capacitors discharge into the primaries of the high voltage trigger transformers thus
providing approximately a 6 KV pulse at the flashtube. As the Xenon
is ionized by this high voltage pulse the flashtubes conduct and emit
a pulse of light. Refer to Chapter 14 for more detailed discussion of
LASCR's and light couplers.
161
SCR MANUAL
00
"0
FIGURE 6.11
TRIGGERING OF SERIES CONNECTED SCR'S WITH LIGHT
K
3K
10K
.
3K
~~.
C/'0'1
FT
i.
-30 ""
~
~-30
_h
5K
470
470
4.70
4.7K
UTe
UTe
PF-7
~
.2,..f
seR~
22.t
--<
50,.
~
~
o.lpf
;:~.
seli
;:;
.'.'
0 *...
t
'OK
27
27
•
162
&
@'pt
Qlpf
10K
FIGURE 6.12
~~
0--
z
CR I ,CR -GE Z4XI48
01 • 02 - GE 2N2646
SCR, • SCRz - GE CI06D
TRIGGER CIRCUIT FOR LIGHT TRIGGERING OF SERIES SCR'S
~ 350
we
SERIES AND PARALLEL OPERATION
6.1.3.3 Slave Triggering for Series SCR's
Slave triggering is a technique for obtaining tum-on of more than
one SCR by applying a trigger signal to only one SCR.8,9 This approach,
although a simple one to implement, has a rather serious limitation.
Rather than simultaneous turn-on, one obtains staggered triggering so
that total turn-on-time can be many times that of a single SCR. After
the first few SCR's of a series string turn-on, the forward blocking voltage intended for the entire string must be supported by those units
which have not yet switched. If the forward voltage to anyone SCR
exceeds its PFV rating, permanent damage to the SCR may result. The
use of shunt capacitors tends to limit the rate of rise of forward voltage
on the later SCR's to switch.
Figure 6.13 illustrates a slave triggering technique. A voltage
equalizing network as previously described is connected across the
cells. Only SCR 1 is directly triggered by the pulse source. The gate of
SCR 2 is triggered by the surge of discharge current from capacitor C 1
when the voltage across SCR 1 decreases abruptly as it switches into
conduction. Since the capacitor-resistor shunts, in conjunction with the
SCR's present a balanced bridge to the zener, the triggering circuit to
the SCR's is essentially insensitive to ordinary cyclical variations and
transients from the supply voltage. In many cases the equalization network, optimized according to the procedure described earlier, will
supply the required gate current to trigger. Rectifiers CR 2 and CRa
can be paralleled with the damping resistors to inhibit triggering from
dv/ dt of the line voltage.
C2
SCR2
RSH
-- --lcR2
RO
1
SOURCE
AND
LOAD
RG
-.--
~
®
®
C,
SCRI
j
RSH
----'1
.'.
CR3:
RO
____~t
C 1 ;:: C2
FIGURE 6.13
SERIES OPERATION OF SCR'S USING SLAVE TRIGGERING
163
SCR MANUAL
The minimum capacitance required to supply sufficient gate current to trigger under all conditions is given by:
10
C 1 ·:2:
V
(6.13)
- RG + GT(max) p. fd
IGT(max)
and
RG = (Vz /2.7) - VGT(max) ohms
(6.14)
IGT(max)
where
Vz nominal zener breakdown voltage of CR1 (volts)
IGT(max)
maximum gate current to trigger under any of the
circuit's operating conditions (milliamps)
VGT(max) = maximum gate voltage to trigger at IGT(max) (volts)
It is necessary that points C and D of Figure 6.13 be as closely
balanced as possible. This is to prevent the How of current in the bridge
due to normal cyclical and transient variations of the supply voltage.
Depending on the direction of an unbalance, a positive gate current to
SCR2 can result from either a falling or rising supply voltage. The
slave triggering technique of Figure 6.13 is expandable to more than
two SCR's in series.
·Figure 6.14 shows another method of slave triggering series connected SCR's. Capacitors C b C 2 ••• C n serve a dual purpose iii this
configuration. First, they provide transient voltage equalization and
secondly they provide the slave triggering current at tum-on. As SCR1
is triggered by the master signal, it begins to discharge capacitor C 1
through the gate of SCR2 thus triggering SCR2 • As SCR2 turns on
capacitor C 2 begins to discharge through the gate of SCRa, and so on.
Resistors Rb R2 ••• Rn_1limit the SCR gate currents. Inductor L
limits the di/dt to SCRn • Figure 6.15 shows what overvoltage can be
expected at each SCR during the tum-on interval when employing
slave triggering. Overvoltage as used here is a percentage of Ep.
=
=
R"
c"
l
I
RII;I
I U~SCR"s
R,
t
OCR.
R,
R.
R,
FIGURE 6.14 •SLAVE TRIGGERING OF SERIES CONNECTED SCR'S
164
SERIES AND PARALLEL OPERATION
I
100
/
80
II
j
60
/
0
0
0
V
V
./
V
SCR POSITION FROM BOTTOM OF STRING
FIGURE 6.15
SCR OVERVOlTAGE AT TURN·ON WHEN SLAVE TRIGGERED VS SCR
POSITION IN STRING
6.1.3.4 The Triggering Pulse
For series operation it is imperative to operate the gate well
beyond the locus of minimum triggering (see Figure 4.13) in order to
obtain tum-on in the minimum possible time. In addition, the pulse
should have a very steep rise (ideally about 100 nanoseconds). The
width of the pulse should be sufficient to insure that the SCR will
latch into conduction under all operating conditions. If anode current
swings momentarily to zero during the conducting cycle, the gate pulse
must be maintained over the entire conduction period. The amplitude
of the gate pulse should be the maximum permissible within the
average and peak gate power dissipation ratings of the SCR.
6.2 PARALLEL OPERATION OF SeR's
When the demands for current handling capability to achieve load
current requirements plus additional margins for overload and reliability purposes exceed the capability of the largest single SCR's
presently available with the desired characteristics, the practice of
paralleling SCR's becomes essential. The main consideration for operating SCR's in parallel is the equalization of forward conduction current
through the parallel paths during both steady and dynamic states.
When paralleling low resistance elements, variation in the· magnetic
:flux linked by the parallel circuit can often be the most significant cause
of unequal current balance. In SCR circuits, this general situation is
further aggravated by any non-uniformity between SCR's forward
165
SCR MANUAL
characteristics. Unequal current sharing can lead to a marked increase
in the junction temperature of SCR's that are carrying a disproportionately large share of the total current. The temperature variation
may further accentuate the characteristic differences, thus resulting
in a thermal runaway condition.
6.2.1 SCR Transient Turn-On Behavior
Without considering external balancing, when paralleling cells
the degree of success one obtains is limited by the degree of control the
device designer is able to achieve on two key turn-on characteristics: 10
delay time, t d , and the minimum turn-on voltage, VOn" The delay time
can be interpreted as the time between application of gate signal and
the actual turn-on of the SCR. The minimum turn~on voltage, sometimes referred to as the "finger" voltage, is the minimum forward anode
voltage at which an SCR can be successfully turned-on with a trigger
signal of sufficient magnitude. Differences in delay times can result in
voltage unbalance during turn-on. Differences in finger voltages may
prohibit the SCR with highest turn-on voltages to trigger. Figure 6.16
shows a hypothetical E-I curve for two SCR's with different finger
voltages.
(a)
FIGURE 6.16
(b)
STATIC SCR GATE TURN·ON BEHAVIOR
Obviously, if SCR 1 is connected directly in parallel with SCR2
having identical characteristics, SCR 2 will never turn-on in applications
:requiring zero voltage triggering. When SCR1 is switched on, the anode
voltage of SCR2 would be that of the on-state voltage of SCRl> and
consequently it will never equal or exceed the minimum required anode
voltage to fire SCR2 even if the width of the trigger pulse is greater
than the delay time of the SCR. (Trigger requirements for parallel
operation will be discussed in Section 6.3.5.)
Therefore, it is essential that in direct paralleling of SCR's, the
forward characteristics of each and every cell must be properly matched.
Figure 6.17 shows the direction and roughly the magnitude of
change in delay time vs gate current and junction temperature.
166
SERIES AND PARALLEL OPERATION
8.0
6.0
\
' "" """"
:i
.
i=
< 10 VOL T5
2.T J =25°C
4.0
0
..'"
\
T
NOTES:
J. OFF STATE VOLTAGE
OPEN CKT. VOLTAGE :;::10 VOLTS
CURRENT RISE TIME '5. lOOns
PULSE WIDTH:: DELAY TIME
2.0
\
~
....
~
;;:
~
0.8
"'" ---
0.6
0.4
r\
~
o
SHORT CIRCUIT GATE CURRENT-AMPERES
(a)
2.0
1.5
0
!i
..
0::
UJ
,.;::
:'lUJ
<>
1.0
'"
"-
""
NOTES,
""',
_.1. OFF STATE VOLTAGE < 10 VOLTS ...............
2. GATE CIRCUIT:
15V OPEN CIRCUIT VOLTAGE
~
0.6 I AMPERE SHORT CKT. CURRENT
100 AS CURRENT RISE TI ME
PULSE WIDTH DELAY TIME
0.8
~
...J
u
,.ii:""
0.4
I-
0.2
-40
0
I
40
80
120
25
JUNCTION TEMPERATURE-oC
(b)
FIGURE 6.17
NORMALIZED DELAY TIME VS GATE CIRCUIT CURRENT & NORMALIZED
DELAY TIME VS JUNCTION TEMPERATURE
167
SCR MANUAL
Delay time has a negative coefficient with gate drive, temperature and
switching voltage which can vary widely between types of SCR's. For
SCR's exhibiting a strong dependence of switching voltage with delay
time, direct paralleling of cells is fraught with danger. If two cells are
connected in direct parallel with different delay times and if the switching voltage is low, the SCR with longer delay time may never be turned
on. With sufficient switching voltage and trigger pulse wider than the
delay time of the slow cell, even if turn-on of both cells is ensured,
the turn-on current mav not be shared as intended. In order to achieve
the proper current sh;ring, certain remedial measures must be carefully undertaken. We shall discuss a few selected methods as listed
below.
1. Direct paralleling using SCR's with matched forward characteristics.
2. External forced current sharing using SCR's with unmatched
forward characteristics.
6.2.2 Direct Paralleling Using SCR's With Unmatched Forward
Characteristics and No Sharing Networks
When forced current sharing is not employed, particular care
must be taken to assure that impedance in series with each individual
parallel path is maintained as nearly equal as possible. Wiring and
connections should be uniform in all respects. The tendency for current
to crowd to the outer branches or paths of a parallel network due to
reactive effects is of particular significance at higher frequencies, and
during the switching interval at the beginning and end of each conduction period. Mutual and self-inductance in series with each parallel
path should be equalized where this phenomenon poses a problem.l1
For direct parallel operation, SCR's with low and uniform turn-on
voltages, closely matched forward E-I characteristics, and short delay
times are desirable. Figures 6.18 and 6.19 illustrate the delay time
and turn-on voltage spread of the C.501 respectively.
168
SERIES AND PARALLEL OPERATION
100
80
40
20
'"z
o
8
10
8
\
~
~
S!
\
"
"
I
W
;:::
a
I
O.1l
3.
lOOns GATE CURRENT
RISE TIME
\.
'"
04
O. I
2. 19" I AMPERE (SHORT CIRCUIT)
1\\"-
i
>-
j
w
I
NOTES;
L TJ=25°C
\
~UM
T~ I-----.,.
C501 SERIES
o
6
8
10
12
OFF STATE VOLTAGE-VOLTS
FIGURE 6.18
DELAY TIME VS DFF·STATE
VOLTAGE
UNITS
MFG
C501 SERIES
GATE CIRCUIT:
2 AMP S.C.
15 VOLT Q.C.
lOOns RISE
50)Js PULSE
T j ·2S·C
o
TURN-ON VOLTAGE
FIGURE 6.19
TURN·ON VOLTAGE
DISTRIBUTION
169
SCR MANUAL
Figure 6.20 is the on-state E-I characteristic curve of the C501.
Similar uniform characteristics exist for the C600, C601 and C602
series.
10,000
8.000
6.000
~I-'-.I----
1------.
..'"
/"
/1/
0-
"'",
::>
u
z
0
'";:!z
z
_--:=--
(TJ =125°C)
I
800
/I
600
400
~!"AX.
Jf
1,000
0
::>
~~
/
'"
U)
~.
-
/)
z>- 2,000
00-
'"
-
~~J~'125·C)~
4,000
~
~,
---e~V'"
t----r-- t--t-
r----------- j-----
Il
1/'
,I
j'!
,
U)
i'i
200
100
-~
r-M-~~5·C)
,
I
1.0
/.5
2.0
3.0
4.0
6.0
8.0
10
INSTANTANEOUS ON-STATE VOLTAGE -vOLTS
fiGURE 6.20
C501 fORWARD CONDUCTION CHARACTERISTICS
Use of factory assembled paralled cell heat exchangers provide
for a high degree of uniformity in thermal and electrical parameters.
A secondary benefit is minimum heat exchanger package size. For
operation at maximum current levels water cooling is recommended.
Determination of derating factors using C501 SCR's will be
undertaken using a design example with the G5 liquid cooled post as
the heat exchanger. It is the basic building block of the G4 and G7
exchangers which allow direct paralleling, water cooling of 3 and 2
SCR's respectively.
Assume we have a three phase waveform. What average current
can two unmatched C501 type SCR's handle? Begin by assuming that
the SCR with the lowest V,£ (call this device SCR 1 ) will be allowed
to run at maximum rated junction temperature and nominal RMS
current. For the type C50l SCR this is 125°C and 850 amperes respectively. For three phase operation we shall assume a rectangular current
waveshape of 0.333 duty cycle. The RMS value for a rectangular waveshape is given by the following relationship
I RMS = I pK yDuty Cycle
For the type C501 SCR in a three phase circuit:
850 = I pK • YO.333
I PK = 1470 amperes
170
SERIES AND PARALLEL OPERATION
2,000
II,~
-
MIN
(TJ~125·C)
-
+7II
I ,BOO
'"'"
'"::;"0:
«
I
I-
z
'"::>
1,600
0:
0:
0
'"~
'"z
I
j
1,400
L
0
'"0
'"z
::>
i'!
z
i'!
'"z
j
/
v
1,200
H
1,000
I
/1
/
I
MAX.
{TJ~~
/j
/
1/"-r
-------
AX (TJ=125°C)
V
,,f
- - -----
-
I
/.~.
/ VI
I II
1.4
1.6
2_0
I.B
2_2
INSTANTANEOUS ON-STATE VOLTAGE -VOLTS
FIGURE 6.21
EXPANDED ONoSTATE CHARACTERISTICS TYPE C501 SCR
During the "on" portion of the cycle, SCR 1 may conduct 1470
amperes, Remember that SCR 1 takes a disproportionate share of the
total current. This is the SCR with the lowest on-state voltage drop at
a given current and temperature among the SCR's of that type connected in parallel. SCR 1 is represented by the curve designated "min
(Tj
125°C)" in Figure 6.21. From this curve we may calculate the
power dissipation in SCR 1 when conducting at maximum allowable
junction temperature.
Pdiss(SCRl) = (I pK) VT (Duty Cycle)
(6.16)
(1470) (1.55) (0.333)
760 watts
The problem now is to find how little current other SCR's in
the cluster will conduct. As we have seen the on-state drop across the
parallel combination must go no higher than that permitted by SCR 1
when conducting maximum RMS current at maximum rated junction
temperature. For our assumed example this was 1.55 volts. Note now
that two maximum on-state voltage drop curves are shown in Figure
125°C and one at T j
25°C. With SCR 1 running
6.21; one at T j
at maximum rated junction temperature it follows that high-on-state
drop units will be running at considerably cooler junction temperatures
since they are conducting less current and hence dissipating less power.
=
=
=
=
=
l71
SCR MANUAL.
It is seen that the lower the-junction_temperature the less the current
conducted at the on-state voltage drop determined by SCRI • Naturally,
use of the Tj
25°C curve will give an ultraconservative estimate of
the mismatch between units. Conversely, use of the T j = 125°C curve
is inappropriate silice it makes _the degree of mismatch look better than
it really maybe. We must estimate at what temperature the junction
of this device is operating and then check our estimate.
=
=
100°C. Find
Assume, as an arbitrary starting point, that T j
the point (Figure 6.21) where V F
1.55 and is 25% of the way
between the T j = 125°C and T j = 25°C. We see that this occurs at
a current of 1060 amps.
First we calculate at what temperature_ we must hold the water
of the heat exchanger on which the SCR's are mounted. This is done
from -a knowledge of the power dissipated in SCR I and the maximum
thermal resistance from junction to water. In order to find the lowest
temperature at which the heat exchanger water must be held, we use
maximum thermal resistance Rau. For SCR type C50l mounted in
the G5 heat exchanger post configuration
RaJA(a",) = .085°C/watt (maximum) .
J - Junction
A - Ambient Water
T j - TA = [PdIS8 (SCRlt] X [SU(S",)]
125°C - T A 760 (0.085)
T A = 125°C - 65°C
= 60°C
We see immediately that the junction temperature of SCR2 will
be higher than 60°C. We are interested in finding the lowesttemperature any type C501 SCR will run.
The power dissipation in SCR2 is
PdiSS(SCR2)
(1060) (1.55) (0.333)
= 548 watts
In order to get the lowest probable junction temperature we must
use minimum thermal resistances. Usually, minimum thermal resistances are not given on specification sheets. For the type C50l SCR
mounted on a G5 post exchanger the following thermal resistance -may
be considered as a minimum value.
SU(S"')
0.08°C/watt
We can now compute the junction temperature of SCR2.
T j - TA PdISS (SCR2) X (SJC(S"') (min»
T j - 60°C 548 X 0.08
44
104°C
Tj
We see now that our first guess of T j = 100°C was a_pretty
good one and no further calculation is necessary. Naturally if our
assumption did not correlate with the answer, a new assumption and
interpolation would be necessary.
We can now formulate a general relationship for the maximum
three phase average current that a parallel group of standard C501
SCR's can handle.
=
=
=
=
=
=
172
=
=
SERIES AND PARAllEL OPERATION
=
1470
+ (np -
1) 1060
(6.17)
3
where
np = number of C501 SCR's in parallel
If it is found that heat exchanger water temperature cannot be
held at 60°C, then full current capability of SCR 1 cannot be realized.
One must start with water temperature and adjust average current in
SCR 1 to maintain junction temperature at + 125°C.
It can be seen that in our example we have derated current 14 %
for parallel operation of two unmatched C501 SCR's. Section 6.2.3
defines "% parallel current derating."
If we had assumed single phase current rather than three-phase,
the approach would be quite similar. Using the peak current in SCRlo
peak power is obtained from the "min (Tj
125°C)" curve. For 180°
haH sine wave conduction, average power dissipation is given by the
following empirical relationship:
Pave = (0.286) P pk
(6.18)
The remainder of the calculation follows that outlined for the
three phase example.
Note that all calculations have been made assuming the SCR's to
be in full conduction. With a constant impedance load, if you are
operating within SCR rating at full conduction, you will remain within
rating as conduction angle is decreased. However, with a variable
impedance load, particularly back EMF loads, required current at
maximum retard angle is used to choose the proper SCR. Assuming
operation at maximum allowable RMS current at 120° conduction
angle. the average power dissipation is about 15% less than that at
180° conduction angle and full RMS rating. When phasing back to
less than 120° conduction angle, average power dissipation decreases
about 12-15% more for every additional 30° of retard down to 30°
conduction angle. These rules-of-thumb are used to determine power
dissipation in the one low-forward-voltage-drop cell for which the
specification sheet rating curves do not apply. For all other cells (highforward-voltage-drop units) the specification sheet curves apply.
IAv(max)
=
6.2.3 Use of SCR's With Matched Forward Characteristics
As we have seen from Figures 6.20 and 6.21, the range of forward
characteristics for a given SCR production line can be quite wide. If
we define percent derating for parallel operation as follows
% Parallel Derating = ( 1 where
n:~M) X
100%
(6.19)
IT = Total required load current through parallel arrangement
1M
= Maximum allowable current for a single cell operating
alone
np = Number of cells in parallel
then we see that in our previous example of Section 6.2.2, 14% derating was required when operating two standard cells in parallel
173
SCR MANUAL
+ 1060 )' X 1000170 = 140170
( 1 _ 1470
2 X 1470
In order to allow for less derating, General Electric supplies C501
type SCR's with matched 'Characteristics. The matching is quantified
by specifying. a maximum on-state voltage spread or band. Figure
'6.22 gives the relationship between bandwidth; in millivolts at 1500
amps and 125°C, T l , and % parallel current derating.
16
/
V
12
/
8
V
4
/'
o
NOTES:
I. FOR USE WITH
G5' POST DESIGN
ON G4,G7 EXCHANGERS.
2. VALID fUR PEAK
CURRENTS IN RANGE OF
4400 TO 2,000 AMPS .
V
I
50
"V
V
/
o
V
IW
100
I
I
200
I
250
ON STATE BAND WIDTH-MILLIVOLTS AT 1,500 AMPS,125°C
FIGURE 6.22
% PARALLEL CURRENT DERATING FACTOR FOR THE C501 SCR, FACTORY MOUNTED
The procedure for designing parallel arrangements with matched
SCR's is quite similar to that outlined in Section 6.22 with the following possible exception. Since little derating usually accompanies the
use of matched cells, all cells are operating near maximum rated junction temperature, As such the T j
125°C forward characteristics can
usually be used exclusively with very little error.
Figures 6.23 and 6.24 show two factory mounted liquid cooled
heat exchanger assemblies available from General Electric with or
without matched cells.
=
174
SERIES AND PARALLEL OPERATION
FIGURE 6.23
FIGURE 6.24
TWO C501'S IN PARALLEL MOUNTED IN A G·7 HEAT EXCHANGER
THREE C501'S IN PARALLEL MOUNTED IN A G-4 HEAT EXCHANGER
175
SCR MANUAL
6.2.4 External Forced Current Sharing
If less than approximately 10% current derating is required when
paralleling SCR's with unmatched forward characteristics, external
forced sharing is required. Returning to our example of Section 6.2.2
we found that SCR1 carried an average current of 490 amperes and
SCR2 carried 353 amperes giving a total capability of 843 amperes.
The maximum average capability of one cell is 490 amperes; it
follows that with varying degrees of forced current sharing, one could
approach about 950 average amperes for two cells in parallel. Let's see
how we go about designing such an arrangement. Figure 6.25 shows
such an arrangement. Let's assume we want to force share just enough
to allow 950 amperes of 3 phase average current.
10
SCRI
~j
500 A(AVE)
1500A (PK)
t t
+ t
VI
V2
t t
(
.. 950 AMPS (AVE)
SCR2 2850AMPS (PK)
l~
450 A lAVE)
1350A IPK)
V4
t
FIGURE 8.25 PARALLEL OPERATION OF SCR'S WITH FORCED SHARING
At % duty factor (three-phase operation) the current through the
pair during the "on" portion of the cycle must be 2850 amperes. With
1500 amperes allowed in SCRlo SCR2 must handle 1350 amperes.
Reading on-voltage from Figure 6.21, the relationship VI + Vs = V2
+ V4 can be solved.
~+~=~+~
~~
1.56 + 1500 Z = 1.71 + 1350 Z
:. Z = 1.0 X 10- 3 Ohms
If resistors are used to effect the sharing, they will indeed be
effective, but necessarily inefficient. In this example, the 1.0 milliohm
resistor in series with SCR1 will dissipate 250 watts.
Current sharing with reactors is a more efficient method than with
resistors.12 Figure 6.26 shows a 1: 1 ratio reactor in bucking connection
for two SCR's in parallel operation.
~.
FIGURE 6.26 EXTERNAL FORCED CURRENT SHARING WITH PARALLEL REACTORS
If the current through SCR1 tends to increase above the current
through SCR2, a counter EMF will be induced proportional to the
176
SERIES AND PARALLEL OPERATION
unbalanced current and tends to reduce the current through SCR 1. At
the same instant, a boosting voltage is induced in series with SCR2
increasing the current flow through the cell. When two matched cells
are used, the magnetic flux balance each other and the core becomes
unnecessary. The most important magnetic requirements of such a
reactor are high saturation and low residual flux densities in order to
provide as great a change in total flux each cycle as possible. An effectively designed balancing reactor will produce a peak voltage equal to
the maximum overvoltage deviation of the two cells throughout the
entire conduction period without saturating. In a single phase circuit
the paralleling reactor should be able to support
~~
volt-second with-
out saturation where
f = supply frequency, Hz
~ V = maximum on-state voltage mismatch between two
SCR's
In a conservative estimation, ~V equals .5 volt at peak load current.
Figure 6.27 and 6.28 illustrate how equalizing reactors can be
used in paralleling odd and even numbers of SCR'S.13 These arrangements may be physically cumbersome and relatively expensive, but they
are highly reliable when continuous operation under partial fault conditions must be provided.
'---f+I'+---+--1-l
I+)--H~..J
FIGURE '.27
THREE SCR'S IN PARALLEL CONNECTION
1+)
FIGURE 6.28
FOUR SCR'S IN PARALLEL CONNECTION
177
SCR MANUAL
6.3 Triggering of Parallel Connected SCR's
If parallel SCR's are triggered from a common source, which is an
essential requirement when switching high currents to large resistive
or capacitive loads, each cell must be supplied with sufficient drive to
exceed its own specific triggering needs. As previously pointed out,
triggering requirements may differ quite widely between individual
SCR's, whether or not units are. parallel matched. As such the suggestions of Section 6.1.3 apply. In addition, it is necessary to drive the
gate hard, regardless of the structure and sensitivity of the gate, commensurate with the peak and average gate power dissipation ratings in
order to insure fast turn-on. This will help the SCRls to share the
switching duty.
At low values of anode current the forward voltage-current characteristic changes from a positive to a negative resistance as current is
reduced to the holding current. Below this value, the SCR will turn
off by reverting to the forward blocking state. This transition point
between positive and negative resistance is represented by the valley
indicated in Figure 6.29, i.e., the current at which minimum forward
voltage drop occurs. It is very difficult to match SCR's satisfactorily
1.0
TO BLOCKING
STATE
o~----------~--------o
2.0
4.0
1.0
3.0
FORWARD VOLTAGE DROP
FIGURE 6.29
ANODE VOLTAGE·CURRENT RELATIONSHIP OF SCR'S WITH
MATCHED FORWARD CHARACTERISTICS
for identical characteristics in this region, particularly over wide
ranges of temperature. This poses no problem when the gate signal
is supplied to the parallel SCR's throughout the anode conduction
period, since any instability in current-sharing or a tendency of one
SCR to turn off will not overheat the other device(s) in parallel because
of the low current level in the region of the valley. As long as gate
current is maintained, an SCR with a tendency to turn off at low
anode current levels will switch into conduction again as soon as
the total load current moves out of this valley area. It will thus assume
its share of the load before overloading on its partner can occur. When
a pulse type of gate signal is employed for triggering paralleled SCR's,
instability may be encountered at low anode current levels which may
have serious consequences if high levels of current follow. Pulsed gate
signals are typical of unijunction transistor triggering circuits and some
types of saturable reactor triggering schemes. Unless the total load
current has reached a sufficiently high level to keep all of the parallel
178
SERIES AND PARAllEL OPERATION
cells above the valley point by the time the gate pulse is removed, a
cells such as A in Figure 6.29 will tum off. In the absence of any further
gate signal, it will remain in the non-conducting state through the
remainder of that cycle, thus failing to carry its share of the load. This
phenomenon is likely to occur when operating at very large conduction
angles in phase controlled AC circuits and when triggering from reactive
lines or into inductive loads where the buildup of load current to
normal levels is restrained by the inductive effect.
For the above reasons use of a maintained gate signal is recommended for triggering parallel SCR's whenever possible. 14
REFERENCES
I. "The Rating of SCR's When Switching Into High Currents,"
Neville Mapham, IEEE Transactions Paper 63-1091.
2. "Overcoming Tum-On Effects in Silicon Controlled Rectifiers,"
Neville Mapham, Electronics, August 17, 1962.
3. "Behavior of Power Semiconductor Devices Under Transient Conditions in Power Circuits," 1. Somos and D. E. Piccone, IEEE
Conference Record, Industrial Static Power Conversion, November
1965.
4. "Switching Characteristics of Power Controlled Rectifiers - TurnOff Action and dv/dt Self Switching," 1. Somos, IEEE Transactions of Communications and Electronics, Vol. 83, No. 75, November 1964.
5. "The Calculation of Tum-On Overvoltages in High Voltage DC
Thyristor Valve," G. Karady and T. Gilsig, IEEE Winter Power
Meeting, February 1971, Paper No. 71 TPI79-PWR.
6. "Application of Thyristors to High Voltage Direct Current Power
Transmission - Sequential Firing," J. L. Hay and K. R. Naik, IEEE
Conference Publication No. 53, Part I - Power Thyristors and
Their Applications, June 1969.
7. "The Light Activated SCR," E. K. Howell, Application Note
200.34*, General Electric Company, Auburn, N. Y.
8. "A 20 KVA DC Switch Employing Slave Control of Series Operated Gate Controlled Switches," J. W. Motto, WESCON CP
64-9.1.
9. "The Gate Cathode Connection for Controlled Rectifier Stacks,"
R. A. Zakarevicius, Proceedings of IEEE, October 1964.
10. "Development in SCR's and Diodes for Power Control Circuits,"
F. B. Golden, Seminar Note 671.15*, General Electric Company,
Auburn, N. Y.
II. "Operation of Unmatched Rectifier Diodes in Parallel Without
ArtifiCial Balancing Means," A. Ludbrook, IEEE CP 63-1169.
12. "Current Balancing Reactors for Semiconductor Rectifiers," 1. K.
Dortort, AIEE TP 59-219.
13. "Parallel Operation of Silicon Rectifier Diodes," G. T. Tobisch,
Mullard Technical Communications, Vol. 8, No. 73, October 1964.
14. "Phase Control of SCR's With Transformer and Other Inductive
AC Loads," F. W. Gutzwiller and J. D. Meng, Application Note
200.31 *, General Electric Company, Auburn, N. Y.
*Refer to Chapter 23 for availability and ordering information.
179
SCR MANUAL
NOTES
180
7
THE TRIAC
7.1 DESCRIPTION
"TRIAC" is an acronym that has been coined to identify the
triode (three-electrode) AC semiconductor switch which is triggered
into conduction by a gate signal in a manner similar to the action of
an SCR. The triac, generically called a Bidirectional Triode Thyristor,
first developed by General Electric (patent No. 3,275,909 and others
applied for), differs from the SCR in that it can conduct in both directions of current How in response to a positive or negative gate signal.
The primary objective underlying development of the triac was to
provide a means for producing improved controls for AC power. The
use of SCR's has proven the technical feasibility and benefits of the
basic functions of solid-state switching and phase-control. In many
cases, however, use of these functions has been limited by cost, size,
complexity, or reliability. The triac development was based upon a
continuing study of various ways for improving overall feasibility of
the basic functions, including evaluation of circuits and components.
To this end, the development appears to have been notably successful,
particularly in the most simple functions.
At this time triacs are available from General Electric in current
ratings up to 25 amperes and voltages to 500 volts. These devices are
available in four different packages as shown in Figure 7.1. Triacs
are rated for both 50-60 Hz and 400 Hz. For abbreviated specifications
of these devices see Chapter 22.
(a) Molded Silicone Plastic, TO·22D. Availalile
in 6 and 10 Amperes.
PRESS·FIT
STUD
(II) Me1a1 Can Versions. Available in 3, 6, 10, 15
and 25 Ampere TYPes.
FIGURE 7.1
TRIAC PACKAGES
181
SCR MANUAL
7.1.1 Main Terminal Characteristics
The basic triac structure is shown in Figure 7.2(a). The region
directly between tenninal MTl and tenninal MT2 is a p-n-p-n switch
in parallel with an n-p-n-p switch. The gate region is a more complex
arrangement which may be, considered to operate in anyone of four
modes: .mrect gate of normal SCR; junction gate of nonnal SCR;
remote gate of complemerrtary SCR with positive gate drive; and remote
gate of complementary SCR with negative gate drive. For more
detailed,,;explanation of triac operation, see Section 7.2.
Figure 7.2 also shows the triac symbol, oriented in proper relationship to the structure diagram; Note that the symbol, although . not
fully definitive, is composed of ,the popularly accepted SCR symbol,
combined with the complementary SCR symbol. Since the terms
"anode" and "cathode" are not applicable to the triac, connections are
simply designated by number. Terminal MTl is the reference point
for measurement of voltages and currents at the gate tenninal and. at
tenninal MT2.
(a)
TERMINAL MT2
"'~P
Jl'HE. ATNSINK
N·
N
P
SILICON
PELLET
MT)
GATE
FIGURE 7.2
TERMINAL MT)
THE TRIAC; (A) SIMPLIFIED PELLET $11IUCTURE. (8) CIRCUIT SYMBOL
The AC volt-ampere characteristic of the triac, Figure 7.3; is
based on tenninal MTl as the reference point. The first quadrant, Q-I,
is the region wherein MT2 is positive with respect to MT1 and vice
versa for Q-III. The breakover voltage, V(BO), in either quadrant (with
no .gate signal) must be higher. than the peak of the nonnal AC wavefonn applied in order to retain control by the gate. A gate current of
specified amplitude of either polarity will trigger the triac into con,duction in either quadrant, provided the applied voltage is less than
V(BO)' If V(BO) is exceeded, even transiently, the triac will switch to
the conducting state and remain conducting until current drops below
the ''holding current," I H • This action provides inherent immunity for
;the triac from excessive transient .voltages and generally eliminates the
need for auxiliary protective devices. In some applications the turning
on of the triac by a transient could have undesirable or hazardous
results on the circuit being controlled, in which case transient suppression is required to prevent turn-on, even though the triac itself is not
damaged by transients.
182
THE TRIAC
I
+
-',~,
I
QUADRANT
I
"l __ ~: =~'
~Q-u4AM~AN=T=m--------l+----=~=+=V(~B-OI--~+V
(MTZ NEGATIVE I
FIGURE 7.3 AC VOlT·AMPERE CHARACTERISTIC OF THE TRIAC
Triac current ratings are based on maximum junction temperature, similar to SCR's. The current rating is determined by conduction
drop, i.e., power dissipation,and thermal resistance junction to case,
and is predicated on proper heatsinking. If the case temperature is
allowed to go above its rated value, as determined from the specification sheet, the triac can no longer be guaranteed to block its rated
voltage, or to reliably turn off when main terminal current goes through
zero. For more details on current ratings of SCR's and triacs, see
Chapter 3. For information on proper heatsink design, see Chapter 18.
For inductive loads, the phase-shift between line current and line
voltage means that at the time that current drops to the IH value and
the triac changes to the non-conducting state, a certain line voltage
exists which must then appear across the triac. If this voltage appears
too rapidly, the triac will immediately resume conduction. In order to
achieve proper commutation with certain inductive loads, the dv/ dt
must be limited by a series RC circuit in parallel with the triac, or current, voltage, phase-shift, or junction temperature reduced. For further
information on the use of triacs with inductive loads, see Section 7.1.4.
7.1.2 Gate Triggering Characteristics
Since the triac may be triggered with low energy positive or negative gate current in both the first and third quadrants, the circuit
designer has a wide latitude for selection of the control means. Triggering can be obtained from DC, rectified AC, AC, or pulse sources such
as unijunction transistors, neon lamps, and switching diodes such as
the ST-2 "diac," the silicon bilateral switch (SBS), and the asymmetrical trigger switch (ST-4).
The triggering modes for the triac are:
MT2+, Gate + ; 1+; First quadrant, positive gate current
and voltage.
MT2+, Gate-; 1-; First quadrant, negative gate current
and voltage.
183
SCR MANUAL
MT2-, Gate +; 111+; Third quadrant, positive gate current
and voltage.
MT2-, Gate-; 111-; Third quadrant, negative gate current
and voltage.
The sensitivity of the triac, at present, is greatest in the I + and
III - modes, slightly lower in the I - mode, and much less sensitive in
the III + mode. The III + mode should not be used, therefore, unless
special circumst-ances dictate it. In such a case, triacs which have been
specially selected for the application are available and should be
specified.
The V-I characteristics of the triac gate shows a low non-linear
impedance between gate and terminal MT 1. The characteristic is similar to a pair of diodes connected in a inverse parallel configuration.
Since in any given mode this characteristic is similiar to an SCR gate,
the gate requirements are rated exactly like SCR's. For details on
gate trigger ratings, see Chapter 4.
7.1.3 Simplified Triac Theory
Four basic thyristor concepts provide a foundation for the theory
of bidirectional thyristor operation. These concepts are:
a) The basic reverse blocking triode thyristor (or SCR)
See Section 1.4.
b) The shorted emitter thyristor
See Section 1.5.
c) Junction gate thyristor
+
GATE
I
I
MAIN
PNI'M - - - ; - STRUCTURE
_
•
ELECTRON FLOW
t
CURRENT FLOW
AUXILIARY PNPN
STRUCTURE
ANODE
FIGURE 7.4 JUNCTION GATE THYRISTOR
Figure 7.4 shows a typical junction gate thyristor structure.
Initially, gate current IG forward biases the gate junction P2-na of
the auxiliary Pl-n1-P2-nS structure, and this structure turns on in
conventional p-n-p-n fashion. As Pl-nl-P2-nS turns on, the voltage
drop across it falls, and the right hand section of region P2 moves
towards anode potential. Since the left hand section of P2 is
clamped to cathode potential, a transverse voltage gradient now
exists across P2, and current flows laterally through P2. As the
right hand edge of P2-n2 becomes forward biased, electrons are
184
THE TRIAC
injected at this point and the main structure turns on (compare
this action to that of the shorted emitter structure).
d) Remote gate thyristor
A remote gate thyristor is one that can be triggered without
an ohmic contact to either of itsintemal base regions. Figure 7.5
depicts a typical remote base structure.
The external gate current IG causes Prns to become forward
biased, and inject electrons as shown. These electrons diffuse
through region PI and are collected by junction Pl-nl' Note that
junction Pl-nl can still act as a collector even though it is forward
biased,4 since the electric field associated with it is in the same
CATHODE
N2
t ELECTRON
FLOW
I
ANODE
FIGURE 7.5
REMOTE GATE THYRISTOR
direction as it would be if PI-ni were reverse biased, as a "collector" normally is. The electrons from ns collected- by Pl-n l cause
an increase of current across PI-n b regeneration starts, and the
structure turns on.
The salient features of the four devices just described can be combined into a single device-the "triac"-which can block voltage in
either direction, conduct current in either direction, and be triggered
on in either direction by positive or negative gate signals. Figure 7.6
is a pictorial view of a typical device. Operation is as follows:
a) Main terminal #2 positive, positive gate current
In this mode the triac behaves strictly like a conventional
thyristor. Active parts are Prnl-P2-n2'
b) Main terminal #2 positive, negative gate current
Operation is analogous to the junction gate thyristor. PI-nr
P2-n2 is the main structure, with na acting as the junction gate
region.
c) Main terminal #2 negative, negative gate current
Remote gate mode. P2-nrPrn4 is the main structure, with
junction P2-na injecting electrons which are collected by the
P2-nl junction.
d) Main terminal #2 negative, positive gate current
P2-n2 is forward biased and injects electrons which are collected by P2-nl' P2-nl becomes more forward biased. Current
through the P2-nl-PI-n4 portion increases and this section switches
on. This mode, too, is also analogous to remote gate operation.
Reference 1 gives a more detailed description of triac triggering.
185
SCR MANUAL
MAIN TERMI.NAL #1
MAIN TERMINAL"'" I
SIDE # I
FIGURE 7.6 TYPICAL TRIAC STRUCTURE
7.1.4 Commutation of Triacs
One important difference between use of a pair of SCR's and
use of a triac in an A-C circuit is that with SCR's each SCR has an
entire half cycle to turn off, while the triac must turn off during the
brief instant while the load current is passing through zero. For resistive
loads this is fairly simple to accomplish since the time available for the
triac to turn-off extends from the time the device current drops below
holding current until the reapplied voltage exceeds the value of line
voltage required to allow latching current. With inductive loads the
task of commutating the triac becomes more difficult.
LOAD
~
COMMUTATION
POINT
SEE FIG.7.S
FIGURE 7.7
INDUCTIVE LOAD WAVEFORMS
Figure 7.7 shows the triac voltage and current waveforms for a
typical inductive load circuit. If we were to examine the waveforms
at the current zero (i.e., at the turn-off point), a waveform such as
Figure 7.8 would be found.
186
THE TRIAC
FIGURE 7.8 TRIAC CURRENT AND VOLTAGE AT COMMUTATION
It can be seen from the current waveform in Figure 7.8 that the
recovery current is acting as a virtual gate current and trying to turn
the device back on. In addition there is a component to the reverse
current which is due to the junction capacitance and the reapplied
dv/dt. This component directly adds to the recovery current but does
not appear until the triac begins to block the opposite polarity.
Section 3.13 discusses the reverse recovery phenomenon in
SCR's. As the rate of removal of current (-di/dt) decreases, the recovery current also decreases. This then implies that at lower dildt's,
higher reapplied dv/dt's are permissible for a given commutation
capability.
An example of such a relationship is shown in Figure 7.9. If the
dv/dt is above this value then additional protection circuits must be
incorporated. The standard method is to use an R-C snubber such as
Rb C 1 in Figure 7.7. The values of Hl and C 1 are a function of the
load, line voltage and triac used. For aid in the choice of Rl and C h
Section 16.3 covers the subject in greater detail.
187
SCRMANUAl
r----------------ID'O
50
NOTES'
I. TJ =115"<:
2.FOR 360"CONDUCTION
di/dt" IT(RMS)w
dj~tlS
~"'-
:g2Q
5
~
WHERE
IN
AMPERES! M ILLISECOND ~
AND IT(RMS) .ISIN
10
AMPERES.
I
g
'" "'"""
I"...
........
~
,~
2
I
2
10
20
50
100
d i !dl- AMPERES! MILLISECOND
FIGURE 7.9
TYPICAL RATE OF REMOVAL OF CURRENT (di/dt) EFFECT UPON
COMMUTATING dv/dt FOR SCBO/B1 TRIACS
7.1.5 Triac Thermal Resistances
R8C
(a) JEDEC Thermal Resistance (b) Triac Effective Thermal Resistance
FIGURE 7.10
THE TWO DIFFERENT TRIAC THERMAL RESISTANCES
On GE triac data sheets two different thennal resistance values
are specified for the same device. This at first sounds impossible, but
consider what these two numbers mean and why they're there.
1) JECEC Thermal Resistance
This thermal resistance specification, usually found in the
Characteristics Table of GE Triac Spec Sheets, is a thennal characteristic specified by JEDEC for purposes of establishing device
interchangeability. It is the value obtained by measuring the peak
junction temperature rise, above the case reference point, produced by a unidirectional DC power being dissipated in the
device. The conduction direction for which this thennal resistance
188
THE TRIAC
value applies is the one that yields the highest value, assuming
that the thermal characteristic is not quite the same for both
conduction directions.
2) Apparent Thermal Resistance
A triac is generally used in AC applications, and consequently,
the JEDEC unidirectional thermal resistance value would yield
a somewhat conservative device AC current rating when using
it in the current maximum case temperature rating calculations.
To overcome this, GE establishes an "apparent" thermal resistance value which when multiplied by the average power, produced by a full sinewave of current of specified frequency, yields
the instantaneous junction temperature at the end of each half
cycle of current conduction. The current rating is so established
that this value of instantaneous junction temperature is the maximum rated value for the device. This assures that the device is
ready to block full rated off-state voltage (within dv/dt limitations) following any half cycle current conduction interval.
This "apparent" thermal resistance of the triac can be represented by a "Y" model as shown in Figure 7.1O(b). The branches
of the Y (ReA' Ren) each represent the thermal resistance of
approximately half of the silicon element (operation for one
polarity of circuit current). The common leg of the Y represents
the thermal resistance of the package hase from the point of silicon
element attachment to the reference point (Tc). GE also estahlishes an apparent transient thermal impedance curve for use in
AC overload current calculations. Again the average power produced by any given number of full cycles of AC current multiplied
by the corresponding value of thermal impedance taken from the
curve will yield the instantaneous junction temperature at the end
of the appropriate half cycle current conduction interval.
7.2 USE OF THE TRIAC
The versatility of the triac and the simplicity of its use make it
ideal for a wide variety of applications involving AC power control.
7.2.1 Static Switching
The use of the triac as a static switch in AC circuits gives many
definite advantages over mechanical switching. It allows the control
of relatively high currents with a very low power control source. Since
the triac "latches" each half cycle, there is no contact bounce. Since
the triac always opens at current zero, there is no arcing or transient
voltage developed due to stored inductive energy in the load or power
lines. In addition, there is a dramatic reduction in component count
compared to other semiconductor static switches.
The most striking example of circuit simplification is seen in the
elementary static switch shown in Figure 7.1l(a). The glass-enclosed
magnetic reed switch provides many million operations from a perma189
SCR MANUAl
nent magnet or from a DC electromagnet "relay" coil. Since the contacts only handle current during the few microseconds required to
trigger the triac, a wide variety of smalt switching elements may be
used in place of the reed switch, such as relays, thermostats, pressure
switches, and program/timer switches. In many cases, snap action of
triggering contacts can be eliminated, thus reducing their cost as well.
This circuit uses gate triggering modes MT2+, Gate+ and MT2-,
Gate-. Figure 7.1l(b) shows the use of a low current diode in series
with the surge limiting resistor, and a three position switch, to obtain
FILAMENT
TRANSFORMER
120V
AC
GE 2DRI5
REED
SWITCH
12,pcV
120V
AC
u
\~II
TRIAC
TI
RI
TRIAC
BASIC STATIC SWITCH
3 POSITION STATIC SWITCH
{al
{bl
FIGURE 7.11
ISOLATED LOW VOLTAGE CONTROL
(e
I
STATIC AC SWITCHING APPLICATIONS OF THE TRIAC
a simple 3 position power control. In position one, there is no gate connection, and the power is off. In position two, gate currenf is alloweo
in one half cycle only, and the power in the load is half-wave. In posi.
tion three, there is gate current for both half cycles, and the power is
on full. As shown in Figure 7.1l(c), the switch can be replaced by a
transformer winding. This circuit makes use of the' difference in primary impedance between the open.circuit and shorted secondary cases.
The resistance R is chosen· to shunt the magnetizing .current of the
primary to ground. This circuit provides control with isolated low
voltage contacts.
Resonant-reed relays have also been used with the triac in the
circuit of Figure 7.1l(a) to provide very sharp frequency-selective
switching in response to coded audio input signals in multi-channel
operations. At the lower frequencies some modulation of triggering
point results from beating with line frequency.
Other useful switching circuits are shown in Figure 7.12, showing
DC and AC triggering for the triac. Switch 8 1 may be replaced by a
transistor which is controlled by a thermistor or a photocell, or other
electrical signal as shown in Figure 7.13. The AC signal of Figure
7.12(b), could be 60 Hz if phased properly to trigger early in each
half cycle of the supply wave. Higher frequencies, above 600 Hz, are
also effective and reduce the size of T, but produce very slight irregularities in triggering point, which are usually negligible~· Frequency
selectivity may be obtained by tuning T or by use of other static or
dynamic filter.circuits for remote-control work or for tape-recorder
programming of'a system. In any case the trigger signal should be
signmcantlyON m' OFF since the trigger sensitivity of the triac is not
quite uniform in both polarities or both quadrants and should not be
used, therefore, as a threshold detector.
190
THE TRIAC
The transistor connections of Figure 7.13 are ideal for driving the
triac, or an array of triacs, from a low' level DC logic source. One
example of this is illustrated by Figure 7.14 which shows two triacs
being driven by a transistor flip-Hop circuit in an AC power Hasher
arrangement.
For further informative details on static switching, see Chapter 8.
AC
LINE
~
TRIAC
TRIAC
SI
_
~
+=-
DC
(b) AC CONTROL
(a) DC CONTROL
FIGURE 7.12
ELECTRICALLY ACTUATED AC STATIC SWITCHES
TRIAC
FIGURE 7.13
TRANSISTOR GATING CONTROL
r
TO
IZOVAC
1
.,
'-MAKE
CCN£CTION
HERE WHEN ONLY
TRIAC I IS NEEDED
CAl
CO2
TI 120:'2.6 STEPDOWN
TRIAC 1- TRIAC 2 : GE seS8 FOR IKW I..OAD
GE SCI4118 FOR 600W LOAD
C2
CR'
eft! - CR4;
GE AI4F
CR5,CR6:
GI[
1N4009
Qf ; GE 2N2646
02,03: GE 2H3416
CI : 500p.F2!5V ELEtTRa..YTIC
Q:o.¥
C3,C4:o.m
AI: MnZW
R2:2 MEG TWlIIIIIER
R3: I MfG
1M: 1000
R&.,M:53Q
R7, ....... :UOQ
10K
RIO,RII.RlZ,"'~:
FIGURE 7.14 A·C POWER FLASHER. TRIACS 1 AND 2 ALTERNATE THEIRONoSTATE
AT A FREQUENCY DETERMINED BY THE SETTING OF R2
7.2.2 Firing With a Trigger Diode
Only four components are required to form the basic full wave
triac phase control circuit shown in Figure 7.15. Adjustable resistor
Rl and capacitor C 1 are a single-element phase-shift network. When
191
SCR MANUAl
the voltage across C1 reaches breakover voltage, V(BO), of the diac,
a bi-directional trigger diode, C 1 is partially discharged by the diac into
the triac gate. This pulse triggers the triac into the conduction mode
for the remainder of that half-cycle. Triggering is in the 1+ and IIImodes in this circuit. Although this circuit has a limited control range,
and a large hysteresis effect at the low-output end of the range, its
unique simplicity makes it suitable for many small-range applications
such as lamp, heater and fan-speed controls.
1--.._-----,--------,
~
~IOO
I
I
TRIAC
I
I
I
I
I
DIAC
GE ST-2
-LO.IILt
'~'(FOR INDUCTIVE
I
1>--------<>--------'---- -- -_,
FIGURE 7.15
LOADS)
BASIC BIAC-TRIAC PHASE CONTROL
To eliminate some of the problems of this basic circuit, more
sophisticated circuits are generally used where the full control range
is required. Other types of bidirectional trigger diodes, such as the
Asymmetrical Trigger Switch (ST-4) may also be used. More details
on this type of firing circuit will be found in Chapter 9.
7.2.3 Other Triggering Methods
In addition to the diac (ST-2) and ATS (ST-4) mentioned above,
devices such as the Unijunction Transistor and Programmable Unijunction Transistor can also be used as triac triggers. Chapters 4 and 9
outline methods for proper circuit designs using these devices.
General Electric also has developed two integrated circuit triggers
for triacs. The first of these is the GEL300, Zero Voltage Switching
Integrated Circuit. With this IC, a triac and four outboard components,
a precision temperature control can be built. This IC and its uses are
covered in depth in Chapters 11 and 12.
The second IC is the GEL301, Integrated Phase Control. This
circuit is designed for high gain, feedback, phase control systems.
A description of this IC, along with several examples of its use are
covered in Section 9.7 and Chapter 10.
7.3 TRIAC CIRCUITRY
In general triac circuitry is the same as that of other thyristors.
Scattered throughout Chapters 8, 9,10,11, 12 and 14 are many examples of triac circuits. In designing triac circuits it is necessary to keep
in mind the unique characteristics of triacs. Below is a short check list
192
THE TRIAC
of those things unique to triacs. These items should be added to those
of Chapter 21 when triacs are used.
1) Commutating-Has adequate arrangement been made to
guarantee commutation?
2) Gate Trigger Modes-Has the system been designed so
that variations in sensitivity between trigger modes will
not affect system performance?
REFERENCES
1. Gentry, Scace and Flowers, "Bidirectional Triode P-N-P-N
Switches," Proceedings of IEEE, April 1965.
2. "Solid State Incandescent Lighting Control," R. W. Fox, Application Note 200.53, General Electric Company, Syracuse, N. Y.*
3. "Using the Triac for Control of AC Power," J. H. Galloway, Application Note 200.35, General Electric Company, Syracuse, N. Y.*
4. J. F. Essom, "Bidirectional Triode Thyristor Applied Voltage Rate
Effect Following Conduction," Proceedings of the IEEE, August
1967, pp. 1312-1317.
5. R. J. Buczynski, "Commutating dv/dt and its Relationship to
Bidirectional Triode Thyristor Operation in Full-Wave AC Power
Control Circuits," IEEE Conference Record of 1967 IGA Second
Annual Meeting, pp. 45-49 .
• Refer to Chapter 23 for avai lability and ordering information.
193
SCR MANUAL
NOTES
194
STATIC SWITCHING CIRCUITS
8
STATIC SWITCHING CIRCUITS
8.1 INTRODUCTION
Since the SCR and the triac are bistable devices, one of their
broad areas of application is in the realm of signal and power switching.
This chapter describes circuits in which these thyristors are used to
pedorm simple switching functions of a general type that might also be
performed non-statically by various mechanical and electromechanical
switches. In these applications the thyristors are used to open or close
a circuit completely, as opposed to applications in which they are used
to control the magnitude of average voltage or energy being delivered
to a load. These. latter types of applications are covered in detail in
succeeding chapters.
Static switching circuits can be divided into two main categories:
AC switching circuits and DC switching circuits. AC circuits, as the
name implies, operate from an AC supply and the reversal of the line
voltage turns a thyristor off. Since most triacs are designed for 50-400
Hz operation, applications at higher frequency would dictate the use
of two SCR's in inverse-parallel connection. The maximum frequency
of operation of SCR's however is limited to approximately 30 K Hz by
the tum-off time requirement of the SCR. Above these frequencies the
SCR's may not recover their blocking ability between successive cycles
of the supply. DC switching circuits on the other hand operate from a
DC (or a rectified and filtered AC) source and an SCR must be turned
off by one of the methods described in Chapter 5. In applications where
the circuit tum-off time is limited, special inverter type SCR's (see
Chapter 22) may be required. These types have tested maximum turnoff time specifications.
8.2 STATIC AC SWITCHES
8.2.1 Simple Triac Circuit and Inverse·Paraliel ("Back·to·Back")
SCR Connection
The circuits of Figure 8.1 provide high speed switching of AC
power loads, and are ideal for applications with a high duty cycle.
They eliminate completely the contact sticking, bounce, and wear associated with conventional electromechanical relays, contactors, etc. As
a substitute for control relays, thyristors can overcome the differential
problem, that is the spread in current or voltage between pickup and
dropout, because thyristors effectively drop out every half-cycle. Also,
providing resistor R is chosen correctly, the circuits are operable over
a much wider voltage range than is a comparable relay. Resistor R is
provided to limit gate current peaks. Its resistance (which can include
195
SCR MANUAL
any "contact" resistance of the control device and load resistance)
should be just greater than the peak supply voltage divided by the
peak gate current rating of the SCR. If R is made too high, the SCR's
may not triggerllt the beginning of each cycle, and "phase contro}"bf
the load will result with consequent loss of load voltage and waveforin
distortion. The control device indicated can be either electrical or
mechanical in nature. Light dependent resistors and light activated
semiconductors,photocouplers (see Chapter 14 where normally open
and normally closed light activated relays are shown), magnetic cores,
and magnetic reed switches are all suitable control elements. ln particular, the use of hermetically sealed reed switches as control elements
in combination with SCR's and triacs offers many advantages. The
reed switch can be actuated by passing AC or DC current through a
small winding around it, or by the proximity of a small magnet. In
either case complete electrical isolation exists between the control signal
input, which may be derived from many sources, and the switched
power output. Long life is assured the SCR or triac/reed switch combination by the minimal volt-ampere switching load placed on the reed
switch by the SCR or triac triggering requirements. The thyristor ratings
determine'the amount of load power that can be switched.
1-.---,-- - - -,
I
I
I
GE
INI692
47 OHMS
l
100",
.n~
THYRECTOR
I
SCR2
I
I
R
41
CONTROL OEVICE
REED SWITCH
(CLOSED
RESISTANCE Rcl
SCR I
I
; ..~~-: O.lILF
FOR
OHMS
GE
INI692
I
INOUCTIVE:
o-_ _ _ _......
L~!!S_ ...J
R~ .,/2·V
lGM
-(R +R I
L
C
(a) Basic Triac Static Switch
FIGURE 8.1
WHERE lGM IS PEAK GATE CURRENT
RATING OF SCR
(b) Inverse-Parallel SCR'S
STATIC AC SWITCHES
For simple static AC switching, the circuit of Figure 8.1(a) has
the advantage over that of Figure 8.1(b) in that it has fewer components. The circuit of Figure 8.1(b), and those circuits to follow using
196
STATIC SWITCHING CIRCUITS
the inverse-parallel SCR configuration, should be kept in mind for
applications where the commercially available triacs cannot handle
severe load requirements such as high frequency, voltage, and current.
For inductive loads an RC snubber circuit, as shown, is required. For
more information on selecting the proper snubber circuit see also
Chapter 15. Triacs are available up to 400 Hz. Above 400 Hz the
circuit shown in Figure 8.1(b) should be used.
8.2.2 Static Switching With Separate Trigger Source
Where DC isolation between control signal input and load is
desired without the use of a mechanical switch (for more details on
light emitters, photosensitive devices, and ·photocouplers, see Chapter
14), or saturable core intermediary, or where a widely varying AC
supply precludes satisfactory triggering of the type shown in Figure 8.1,
a triac or a back-to-back pair of SCR's may be triggered from a separate source as shown in Figure 8.2. Here, the high frequency output
of a transistor blocking oscillator, or a UJT free-running oscillator is
transformer coupled to the triac or SCR gates. Suitable oscillator circuits are discussed in Section 4.14. For minimum load waveform distortion and minimum generated RFI, oscillator frequency should be high
enough to ensure that the triac or SCR's trigger early in the AC cycle.
Other types of UJT trigger circuits suitable for use with AC static
switching arrays are described in Chapter 4.
i
AC
SUPPLY
1
{
,,
:
o.l,.F_~_
FOR
INDUCTIVE
LOADS
'i"
,
TRIAC
I
l
loon .....
II2W i'
,
R2, and C l . The upper limit of time delay which can be achieved
depends on the required accuracy, the peak point current of the UJT,
the maximum ambient temperature, and the leakage current of the
capacitor and UJT (lEO) at the maximum ambient temperature. The
absolute upper limit for the resistance Rl + R2 is determined by the
requirement that the current to the emitter of the UJT be large enough
to permit it to trigger (i.e., be greater than the peak point current) or
+R <
R
1
2
(1 -.,,) VI
25 I
-_P+I.,
(8.6)
VI
where ." is the maximum value of intrinsic standoff ratio, VIis the
minimum supply voltage on the UJT, Ip is the maximum peak point
current measured at an interbase voltage of 25 volts, and Ie is the
maximum leakage current of the capacitor at a voltage of ."Vl . If high
values of capacitance are required it is desirable to use stable, low
leakage types of tantalytic capacitors. If tantalytic or electrolytic
capacitors are used it is necessary to consider forming effects which
will cause the effective capacitance and hence the period to change as
a function of the voltage history of the capacitor. These effects can be
reduced by applying a low bias voltage to the capacitor in the standby
condition.
The resistor Ra can serve as a temperature compensation for the
circuit, increasing the value of Rs causes the time delay interval to
have a more positive temperature coefficient. The over-all temperature
coefficient can be set exactly zero at any given temperature by careful
adjustment of Ra. However, ideal compensation is not possible over a
wide range because of the nonlinear effects involved. To reset the
circuit in preparation for another timing cycle SCR I must be turned off
either by momentarily shorting it with a switch contact or by opening
the DC supply.
8.10.2 AC Powered Time Delay Relay
Figure 8.24 illustrates a time delay circuit using a relay output
with a push button initiation of the timing sequence. In the quiescent
state SCR I is on and relay SI is energized. Contact SIA is closed, shorting out the timing capacitor Ca. To initiate the timing cycle push button
switch SW2 is momentarily closed which shorts SCR I through contact
SIB causing SCRI to tum off. When SW2 is released SI is de-energized
and the timing sequence begins. The particular configuration of SW2
and SIB is used to prevent improper operation in case SW2 is closed
again during the timing cycle. Capacitor C s is charged through R5 and
RIO until the voltage across C a rf'Alches the peak point voltage of Ql
causing Ql to trigger. The positive pulse generated across R12 triggers
SCR I which pulls in the relay and ends the timing cycle. The timing
cycle can be terminated at any time by push button switch SWa which
causes current to How in R 13 thus triggering SCRI • Capacitor C 4 supplies current through RIa during the instant after the supply is turned
on thus triggering SCR I and setting the circuit in the proper initial state.
216
STATIC SWITCHING CIRCUITS
RI
R2
CRI CR2
TI
+
60CPS
Ils011
CI
+
C2
CR3 CR4
R I - 211, I WATT
R2, R3 - 33011,1/2 WATT
R4 - 3SIl, SWATT
RS - 2.S K, LINEAR POT
R6 - 2SK, 1/2 WATT
R7 -100K,1/2%, 1/2 WATT
R8 -200K, 1/2%,1/2WATT
R9 -1011,1/2 WATT
RIO - lOOK. 10 TURN HELl POT
RII - ISO.n, 1/2 WATT
RI2-·18Il, 1/2 WATT
R13- 1.2K, 2 WATT
RI4 - 100A, 1/2 WATT
CI- SOOI'FD, SOV
C2 - 1001' FD, SOV
C3 - 100 I' FD, 20V TANTALUM
C4 - 10!,FD, SOV
SCRI - GE CISF
OR CIIF OR CI22
CRI-CR6-GE AI4A
CR7- 18V, 10% 1 WATT ZENER
QI - GE 2NI671B
SI - GE CR2791GI22A4
4PDT RELAY
PLI, PL2 - GE 1447, 24V LAMP
TI-IISV/2SV IA TRANSFORMER
FIGURE 8.24 VARIABLE TIME CONTROL CIRCUIT
The timing interval is determined by the setting of a preclSlon
ten tum Helipot RIo which may be set from 0.25 to 10.25 seconds in
increments of 0.01 second. The initial setting of 0.25 seconds takes
into account the added series resistance of the time calibration potentiometer Rs. Additional series resistance of lOOK and 200K may be
added by SW1 to extend the time range by 10 seconds and 20 seconds.
A fourth position of SWi open circuits the timing resistors and thus
permits unrestricted on-off control of the circuit.
Tests of the circuit have shown an absolute accuracy of 0.5% after
initial'calibration and a repeatability of 0.05% or better.
8.10.3 Ultra-Precise Long Time Delay Relay·
Predictable time delays from as low as 0.3 milliseconds to over
560 OHMS
r-----_4r_----~----_4r_----~--------_.~AA~__O+28V
150
OHMS
GE AI4A
OUTPUT
Q2 GE
2N2646
SCRI
GE CI22F, CI5F
OR GE CIIF
18
VOLT
ZENER
GE
C4
Z4XLI8
_05pf
FIGURE •• 25
ULTRA-PRECISE LONG PERIOD TIME DELAY
217
SCR MANUAL
3 minutes are obtainable from the circuit of Figure 8.25, without
resorting to a large value electrolytic-type timing capacitor. Instead, a
stable low leakage paper or mylar capacitor is used and the peak point
current of the timing UJT (Ql) is effectively reduced, so that a large
value emitter resistor (R I ) may be substituted. The peak point requirement of Ql is lowered up to 1000 times, by pulsing its upper base with
a % volt negative pulse derived from. free-running oscillator Q2' This
pulse momentarily drops the peak point voltage of Ql> allowing peak
point current to be supplied from C 1 rather than via R I . Pulse rate of
Q2 is not critical, but it should have a period T that is less than .02
(Rl . C 1 ). With Rl
2000 megohms and C I
2 pi (mylar), the circuit
has given stable time delays of over one hour. R2 is selected for best
stabilization of the triggering point over the required temperature
range. Because the input impedance of the 2N494C UJT is greater than
1500 megohms before it is triggered, the maximum time delay that
can be achieved is limited mainly by the leakage characteristics of C I .
=
8.10.4
=
Time Delay Circuits Utilizing the Programmable
Unijunction Transistor (PUn
Very simple and precise time delay circuits can be achieved with
the PUT.s Among the important advantages are elimination of calibration pots, longer time delays and low cost.
8.10.4.1 30 Second Timer
Figure 8.26(a) shows a 30 second time delay using the 2N6028.
Here we are taking advantage of high sensitivity to achieve high values
of timing resistance (30 megs). Calibration has been eliminated by
using 1 % components for the intrinsic standoff resistors and also for
the RC timing components. Note the additional use of the compensation diode IN4148.
+~~A~~r-________~__~_
~~AA~T~________~__~_
30M.
1%
30 MEG
±I%
2N6027
If
SCR
~2%
16k
1%
10k
1.0
pF
~2%
c- GEAAIBAIOec
(a)
FIGURE 8.26
(b)
PRE'CALIBRATED 30 SECOND TIMERS
The same performance can be achieved using the 2N6027, as in
Figure 8.26(b). When using the 2N6027, one must either significantly
decrease the value of the timing resistor and increase the capacitance,
or use a sampling scheme as shown here. This is precisely the same
timer as in Figure 8.26(a) with the addition of the 10K resistor, a diode
218
STATIC SWITCHING CIRCUITS
and the sampling transistor. A 1KHz pulse train is applied to the base
of the NPN transistor. Each pulse lasts 10 p.secs. This modulates the
intrinsic standoff voltage once every 1 millisecond to "take a look" at
the capacitor voltage. The 2N6027 derives its peak point current from
the capacitor.
Calibration of timers is easier using the PUT. When RT and C T
are not 1 % parts, the scheme in Figure 8.27 can be used. Here an
inexpensive, low resistance, trimpot can be used instead of a wirewound pot for the time adjustment.
~-;:;;'rnH-l4_~TRIM
POT
2.7
MEG
FIGURE 8.27
CALIBRATION VIA THE "INTRINSIC STAND·OFF RATIO"
8.10.4.2 Long Delay Timer Using PUT
Figure 8.28 shows the use of the PUT's as both a timing element
and sampling oscillator. A low leakage film capacitor is required for C 2
due to the low current supplied to it.
START
+28Vcr--/
15K
1000M
1M
Q2
2N6027
1M
200
QI
CI
5,.F
2N6027
FIGURE 8.28
301(
C2
5,.F
IN4443
100
PULSE
OUTPUT
LONG DELAY TIMER
8JO.5 ASO-Second Time Delay Circuit Switching AC
Figure 8.29 shows a time delay circuit using the triac latching
technique. When capacitor C 1 charges to the breakover voltage of the
diac, the triac triggers and energizes the load. The time delay is determined by the time constant of (Rl + R 2) and Cl' To reset the circuit,
capacitor C 1 is discharged through R3 and Sl'
219
SCR MANUAL
120 VAC
ADJUST
TIME
DELAY
MT2
GE
SC46BI3
TRIAC
R2
120V
60-
R4
100
n
100,.FD
50V
R6
15K
FIGURE 8.29
LOAD
UP TO
1200
WATTS
A SIXTY-SECOND TRIAC TIME DELAY CIRCUtT
8.10.6 One Second Delay Static Turn~Off Switch
An AC-switch with delayed tum-off is shown in Figure 8.30. The
components CRl> Rl> CR 2 and C1 supply about -20 V between the
MT1 and the gate terminal of the triac, but gate current can only How
to trigger Ql when SW is closed to forward bias Q2 into conduction.
As long as SW is closed Q2 is saturated and Ql maintains the load
activated,
lOA
120V
60Hz
CRI
AI4D
CR2
20K ,
IW
RI
IOKn
2W
FIGURE 8.30
220
ONE SECOND DELAY STATIC TURN-OFF SWITCH
STATIC SWITCHING CIRCUITS
When SW is opened C 2 will discharge through R2 and the base
emitter junction of Q2, keeping Q2 and Ql conducting. As C 2 discharges, Q2 will finally tum off and the triac will commutate at the
next zero crossing, interrupting the load current.
Changing the time constant C 2 R2 permits the selection of various
tum-off delay intervals.
8.11 NANOAMPERE SENSING CIRCUIT WITH
100 MEGOHM INPUT IMPEDANCE
The circuit of Figure 8.31 may be used as a sensitive current
detector, or as a voltage detector having high input impedance. A sampling technique similar to that described in the previous section is used
to give an input current sensitivity (lIN) of less than 35 nanoamperes.
Input impedance is better than 100 megohms.
INPUT
TERMINALS
{+
_
R2
100M
+28V~~----1-----~------~--~------~------'
150
OHMS
IK
Q2 GE
CRI
GE
IN 3604
2N2646
SCRI
GE
CI5F
OR C22F
OR CI22
C2
.Olpf
FIGURE 8.31
NANOAMPERE SENSING CIRCUIT
Current gain between output and input of the circuit as shown is
greater than (200 X 10- 6 ).
Resistor RI is adjusted so that the circuit will not trigger in the
absence of the current input signal lIN. lIN then charges C 2 through
the 100 megohm input resistor R2 towards the emitter triggering voltage of Q1. R2, however, cannot supply the peak point current (2 pA)
necessary to trigger Ql> and this current is obtained from C 2 itself by
dropping Q]'s triggering voltage momentarily below VC2. Relaxation
oscillator Q2 supplies a series of .75 volt negative pulses to base two
of QI for this purpose. The period of oscillation of Q2 is not critical
but should be less than .02 times the period of Ql. Capacitor C 2 can
be kept small for fast response time because C l stores the energy
required to trigger SCR I . Rapid recovery is possible because both
capacitors are charged initially from R I . Some temperature compensation is provided by the leakage current of CR I subtracting from the
221
SCR MANUAL
leakage current of Ql. Further compensation is obtainable by adjusting
the value of Ra. A Hoating power supply for the UJT trigger circuit
with pulse transformer coupling from Ql to SCRb enables one of the
two input terminals to be grounded, where this may be desirable.
8.12 MISCELLANEOUS SWITCHING CIRCUITS USING
GE LOW CURRENT SCR'S
The C103 C5, C6 and C106 series of SCR's have a high gate
sensitivity. Gate triggering can therefore be achieved from such low
level elements as thermistors and light sensitive resistors. When used
as a gate amplifier for the higher rated SCR's, either of these devices
makes possible a multitude of solid state thyratron tube analogues.
The C5 SCR is also suitable for use as a very high voltage remote-base
. transistor. For more detailed application "information on the low current SCR's, see Reference 5.
8.12.1 Dual Output, Over-Under Jemperature Monitor
The circuit of Figure 8.32 is ideal for use as an over-under temperature mGmit.or, where its dual output feature can be used to drive
"HICH" and "LOW" temperature indicator lamps, relays, etc.
•
C?----...--..::..II
•
FOil INDUCTIVE LOADS
CR2
GE AI4A
I
I
r--.-'
LOAD ... 2
TI
PRIMARY
CR3
GE AI4A
CRI
GE AI4A
I
NOTES: (I) TI: 6.3 FILAMENT
TRANSFORMER
L
_-...J
I
(2) T: GE 20052 THERMISTOR
FIGURE 8.32
TI
SECONDARY
TEMPERATURE MONITOR
T 1 is a 6.3 volt filament transformer whose secondary winding is
connected inside a four arm bridge. When the bridge is balanced, its
AC output is zero, and the C5 (or C7) receives no gate signal. The
bridge's DC resistance is sufficiently low to stabilize the SCR during
forward blocking periods. * If now the bridge is unbalanced by raising
or lowering the thermistor's ambient temperature, an AC voltage will
appear across the SCR's gate cathode terminals. Depending in which
sense the bridge is unbalanced, positive gate voltage will be in phase
with, or 180 0 out of phase with the AC supply. If positive gate voltage
is in phase, SCR will deliver load current through diode CRl to load (1),
diode CR 2 blocking current to load (2). Conversely, if positive gate
'See Section 4.3.6 "Negative Gate Biasing."
222
STATIC SWITCHING CIRCUITS
voltage is 180° out of phase, diode CR2 will conduct and deliver power
to load (2), CR 1 being reverse biased under these conditions. CRa prevents excessive negative voltage from appearing across the SCR's
gate/cathode terminals. With component values shown, the circuit will
respond to changes in temperature of approximately I-2°C. Substitution of other variable-resistance sensors, such as cadmium sulphide
light dependent resistors (LDR) or strain gauge elements, for the thermistor shown is of course permissible. The balanced bridge concept of
Figure 8.31 may also be used to trigger conventional SCR-series load
combinations. As a low power temperature controller for instance, a C5
could be used to switch a heater element, with a thermistor providing
temperature feedback information to the trigger bridge.
For more information on temperature controls see Chapter 12 on
zero voltage switching.
8.12.2 Mercury Thermostat/SCR Heater Control
The mercury-in-glass thermostat is an extremely sensitive measuring instrument, capable of sensing changes in temperature as small as
O.l°C. Its major limitation lies in its very low current handling capability - for reliability and long life, contact current should be held
below 1 rnA. In the circuit of Figure 8.33 the General Electric C5B or
C106B SCR serve as both current amplifier for the Hg thermostat and
as the main load switching element.
100 WATT HEATER LOAD
GE C5B
OR
CIOSB
120 VAC
SO CPS
GE AI4B
CRI - CR4
TWIST LEADS TO MINIMIZE
PICKUP
HG IN GLASS THERMOSTAT
(SUCH AS VAP AIR DIV. 206-44
SERIES; PRINCO#TI4I, OR
EQUIVALENT)
FIGURE 8.33
TEMPERATURE CONTROLLER
With the thermostat open, the SCR will trigger each half cycle
and deliver power to the heater load. When the thermostat closes, the
SCR can no longer trigger, and the heater shuts off. Maximum current
through the thermostat in the closed position is less than 250 pA rms.
223
SCR MANUAL
8.12.3 Touch Switch or Proximity Detector
The circuit shown in Figure 8.34 is actuated by an increase in
capacitance between a sensing electrode and the ground side of the
line. The sensitivity can be adjusted to switch when a human body is
within inches of the insulated plate used as the sensing electrode. Thus,
this circuit can be used as an electrically-isolated touch switch, or as a
proximity detector in alarm circuits.
47K
10M.
LOAD
1
1M
115 VOLTS
60Hz
SCR
GE
CI06B
1M
j
DIAC
GE
ST2
GE
2N6027
TO
SENSING
ELECTRODE
IK
ALL RESISTORS 1/4 WATT
FIGURE 8.34 TOUCH SWITCH OR PROXIMITY DETECTOR
The GE 2N6027 Programmable Unijunction Transistor (PUT),
will switch "ON" when the anode voltage exceeds the gate voltage by
ail amount known as the trigger voltage (approximately 0.5 volts). This
anode voltage is clamped at the "ON" voltage of the diac (ST2). As
the capacitance between the sensing electrode and ground increases
(due to an approaching body), the angle of phase lag between the
anode and gate voltages of the D13T increases until the voltage differential at some time is large enough to trigger this PUT. Because the
anode voltage is clamped, it is larger only at the beginning of the cycle;
hence, switching must occur early in the cycle, minimizing RFI.
Sensitivity is adjusted with the I megohm potentiometer which
determines the anode voltage level prior to clamping. This sensitivity
will be proportional to the area of the surfaces opposing each other.
8.12.4 Voltage Sensing Circuit
A low cost voltage or threshold detector is shown in Figure 8.35.
224
STATIC SWITCHING CIRCUITS
+25V FULLWAVE
RECTIFIED DC
R2
75
2W
LI
GE
NO. 382
CRI
AI4F
3
RI
10K
CR5
AI4F
SCRI
C7
R4
75
lOW
CR2
AI4F
L2
GE
NO.382
R3
10K
CI
10l£F
50V
CR3
AI4F
CR4
AI4F
4
FIGURE B.35
2
LOW COST VOLTAGE DETECTOR
When +25 V DC (full wave rectified) is applied between terminal
#1 and 2, current will How through R4 , L 2, CRa and CR4 • L2 will be
illuminated and Ll will be dark. As soon as a voltage is applied between
terminals #3 and 4 (3 positive) and a threshold of about 2.8 V is
exceeded SCR1 will turn on, actuating L 1 . L2 will be turned off because
of CR5 •
The threshold voltage can be increased by adding more diodes to
CR h CRa and CR 4 or replacing them by a zener diode.
This circuit is useful in detecting the voltage across an SCR in
the on or off positions or indicating the output state of an operational
amplifier, etc .
.8.12.5 Single Source Emergency Lighting System
An emergency lighting system which maintains a 6 volt battery
at full charge and switches automatically from the AC supply to the
battery is shown in Figure 8.36.
225
SCR MANUAL
CRI
AI4F
r
RI~
SELECT TO GIVE DESIRED CHARGE
RATE (VALUE AND WATTAGE)
SCRI
CI06YI
CR2
AI4F
AD
6.3V
INPUT
50-60
HERTZ
6.3V
CR3
AI4F
6V
LAMP
TI
+
-=6VOLT
-=- BATTERY
FIGURE 8.36 SINGLE SOURCE EMERGENCY LIGHTING SYSTEM
Transformer T 1 and diodes CR 2 and CRa supply DC voltage for
the 6 V lamp load. CR l and Rl supply the battery with charging current, which can be varied by R l . The anode and gate of SCRl are kept
at the battery voltage, while the cathode of SCR l is kept at a higher
potential by C l . Should the voltage on the cathode of SCRl fall below
the battery voltage due to interruption of the AC input, SCRl will
trigger and supply the lamp with power from the battery. When the
AC reappears, SCR l will tum off automatically and the battery will
re-recharge.
8.12.6 Liquid Level Control
When it is desirable to keep the fluid level of a liquid between
two fixed points, this hybrid control is extremely useful. The control
takes advantage of the best characteristics of both power semiconductors and electromechanical devices.
Two modes, for filling or emptying are possible by simply reversing the contact connections of Kl as shown in Figure 8.37.
F,
~--------------
,),.
oR~g9
FIGURE 8.37
226
LIQUID LEVEL CONTROL
STATIC SWITCHING CIRCUITS
The loads can be either electric motors or solenoid operated valves,
operating from AC power. Liquid level detection is accomplished by
two metal probes, one measuring the high level and the other the low
level.
The relay Kl is energized by Ql which is controlled by Q2, a PUT,
whose gate form the detector. The PUT is normally off but when liquid
rises to the high probe level, the impedance of the liquid creates a voltage divider and the PUT triggers. When the PUT conducts it turns on
Ql which will pick up K1. Kl will turn on Qs activating the load and
will also arm the low level probe which holds the circuit on until the
liquid level drops below this probe. At this time the circuit is de-energized, turning off the load.
An inversion of the logic (keeping the container filled) can be
accomplished by replacing the normally open contact on the gate of
Qa with a normally closed contact.
8.13 THYRATRON REPLACEMENT
A thyratron tube is characterized by a very high signal input
impedance, low pick-up and drop-out currents, and good power handling capabilities. On the other hand, it is fragile, requires filament
power, is frequency limited by a long deionization time, and has a
fairly high forward drop. While the solid-state equivalents to this
device, using the C5 as a trigger element for a larger size SCR, can
match the thyratron in input impedance, current handling ability and
low pick-up current, they possess none of the gas tube's limitations.
At the present time, however, the maximum forward blocking voltage
attainable using a single C5 is 400 volts. This can be increased by
series connecting additional SCR's (see Chapter 6).
ANODE
r---------------,
I
CRI
I
(ioon)
I
I
R2
"GRID" I
I
G I
RI
(10K)
GRID-CATHODE VOLTAGE
+
o
I
~
TIME
I
I
IL
__________ _
CRI--GE AI4D
la) Equivalent Circuit
(II) Grid Voltage Waveforms
FIGURE 8.38 SIMPLE THYRATRON REPLACEMENT
Referring to Figure 8.38; with a negative potential on grid terminal "G", stabilizing gate bias is provided through Rl and R3 for the
C5. When the "Grid" is driven positive, however, a maximum current
of 200 microamps will trigger the smaller SCR into conduction. The
C35D is triggered in tum by the C5, and can conduct up to 25 amps
227
SCR MANUAL
rms load current. With the voltage grades shown, the '~device" is capable of blocking voltages up to 400 volts. Over-all pick-up current is
determined by the C5 rather than by the C35, a useful feature when
the "thyratron" is operating into a highly inductive load. Diode CR 1
prevents transistor action in the C5 if positive grid voltage should coincide with negative anode potential.
General Electric is manufacturing the S26 and S27 solid state
thyratrons which are 200 volt devices, but higher voltage devices such
as the SL-3 and SL-4 and custom designs are available. (For more
information on solid state thyratrons see also Reference 7.)
8.14 SWITCHING CIRCUITS USING THE C5 OR Cl06
SCR AS A REMOTE·BASE TRANSISTOR
8.14.1 "Nixie"® and Neon Tube Driver
The C5 SCR, when biased as a remote-base transistor (for detailed
information on remote-base transistors see Chapter 1), makes an excellent high voltage transistor suitable for driving Nixie, neon and other
type of high voltage digital readout displays. Collector voltage rating
of the equivalent transistor equals or exceeds the VBH(FX) --------------~
FIGURE 8.39
"BASE
(C5 CATHOOE)
TRANSISTORIZED NIXIE® DRIVER
8.14.2 Electroluminescent Panel Driver
Either of the circuits of Figure 8.40 may be used to drive the
elements of an electroluminescent display panel, depending on the input
228
STATIC SWITCHING CIRCUITS
logic required. Here, the high voltage capabilities of the C5 SCR are
again combined with its usefulness as a transistor, in this case a
symmetrical transistor, to control full-wave AC drive at high voltage
and frequency, low current.
BLEEDER
MAY BE
REQUIRED
El
ELEMENT
GE
C5D XI42
3V
AC SUPPLY 230V RMS
60 TO 2K HERTZ
-~
= SWITCH
+
(a) Series Drive -
No Signal. Display "Off"
R5
AC SUPPLY 230V RMS
60 TO 2K HERTZ
EL
(b) Shunt Drive FIGURE 8.40
No Signal. Display "On"
ELECTROLUMINESCENT PANEL DRIVER
REFERENCES
1. "Using the Triac for Control of AC Power," J. H. Galloway, General
Electric Company, Auburn, N. Y., Application Note 200.35.*
2. "Solid State Electric Heating Controls," R. W. Fox and R. E. Locher,
General Electric Company, Auburn, N. Y., Application Note
200.58.*
3. "Regulated Battery Chargers Using the Silicon Controlled Rctifier,"
D. R. Grafham, General Electric Company, Auburn, N. Y., Application Note 200.33*
4. "Flashers, Ring Counters and Chasers," R. W. Fox, General Electric
Company, Auburn, N. Y., Application Note 200.48.*
5. "Using Low Current SCR's," D. R. Grafham, General Electric Company, Auburn, N. Y., Application Note 200.19.*
6. "The D13T - A Programmable Unijunction Transistor Types
2N6027 and 2N6028," W. R. Spofford, Jr., General Electric Company, Syracuse, N. Y., Application Note 90.70.*
7. "The Solid State Thyratron," R. R. Rottier, General Electric Company, Auburn, N. Y., Application Note 200.36.*
"Refer to Chapter 23 for availability and ordering information.
229
SCR MANUAL
NOTES
230
9
AC PHASE CONTROL
AC PHASE CONTROL
9.1 PRINCIPLE OF PHASE CONTROL
"Phase Control" is the process of rapid ON-OFF switching which
connects an AC supply to a load for a controlled fraction of each
cycle. This is a highly efficient means of controlling the average power
to loads such as lamps, heaters, motors, DC suppliers, etc. Control is
accomplished by governing the phase angle of the AC wave at which
the thyristor is triggered. The thyristor will then conduct for the
remainder of that half-cycle.
There are many forms .of phase control possible with the thyristor,
as shown in Figure 9.1. The simplest form is the half-wave control of
Figure 9.1(a) which uses one SCR for control of current flow in one
direction only. This circuit is used for loads which require power con-
RECTIFIER
-+\:rAC
SWITCH
CONTROLLED HALF-WAVE
CONTROLLED HALF PLUS
FIXED HALF-WAVE
(a)
(bl
CONTROLLED FULL WAVE
CONTROLLED FULL WAVE
(el
(dl
BRIDGE RECTIFIER
.,
:-DC LOAD
I
..J
TRIAC
ZJ-257B
~
AC
CONTROLLED FULL WAVE
(el
FIGURE 9.1
CONTROLLED FULL WAVE FOR AC OR DC
(Il
BASIC FORMS OF AC PHASE CONTROL
231
SCR MANUAL
trol from zero to one-haH of full-wave maximum and which also permit
(or require) direct current. The addition of one rectifier, Figure 9.1(b),
provides a fixed half-cycle of power which shifts the power control
range to half-power minimum and full-power maximum but with a
strong DC component. The use of two SCR's, Figure 9.1(c), controls
from zero to full-power and requires isolated gate signals, either as two
control circuits or pulse-transformer coupling from a single control.
Equal triggering angles of the two SCR's produce a symmetrical output
wave with no DC component. Reversible half-wave DC output is
obtained by controlling symmetry of triggering angle.
An alternate form of full-wave control is shown in Figure 9.1(d).
This circuit has the advantage of a common cathode and gate connection for the two SCR's. While the two rectifiers prevent reverse voltage
from appearing across the SCR's, they reduce circuit efficiency by their
added power loss during conduction.
The most flexible circuit, Figure 9.1(e), uses one SCR inside a
bridge rectifier and may be used for control of either AC or full-wave
rectified DC. Losses in the rectifiers, however, make this the least efficient circuit form, and commutation is sometimes a problem (see Section 9.3). On the other hand, using one SCR on both halves of the AC
wave is a more efficient utilization of SCR capacity, hence the choice
of circuit form is based on economic factors as well as performance
requirements.
By far the most simple, efficient and reliable method of controlling
AC power is the use of the bidirectional triode thyristor, the triac, as
shown in Figure 9.1(f). Triac characteristics are discussed in Chapter 7.
The fact that the triac is controlled in both directions by one gate and
is self-protecting against damage by high-voltage transients has made
it the leading contender for 120 and 240 VAC power control up to
6 KW at this writing.
9.2 ANALYSIS OF PHASE CONTROL
Rectifiers and SCR's are rated in terms of average current since
this is easily found by a DC ammeter. Most AC loads are more concerned, however, with the RMS, or effective, current, which is why
triacs are rated in terms of RMS current.
Figures 9.2 and 9.3 show the relationships as a function of phaseangle, IX, at tum-on, of average, RMS, and peak voltages as well as
power in a resistive load. Since the SCR is a switch it will apply this
voltage to the load, but the value of current will depend on load
impedance.
As an example of the use of these chmts, suppose it is desired to
operate a 1200 watt resistive load, rated at 120 volts, from a 240 volt
supply. Connection of this load directly to the supply would result in
4800 watts, therefore the desired operation is at % maximum power
capability. We may use, for this case, either a half-wave or full-wave
form of control circuit.
Starting with the half-wave case and Figure 9.2, the % power
point is a triggering phase angle of 90 degrees. Peak output voltage
Epo is equal to peak input voltage, 1.41 X 240 volts or 340 volts.
232
AC PHASE CONTROL
"CONDUCTION ANGLE" =.leo-,. (I
o
(I
1.0
......
,L.Epo
-
Ep
.9
-"
.8
--
.7
1\
.6
0:
~
~
.5
"",--
--,...,
ERMa
Ep
"
.4
FULL-WAVE (R
LOAD}p~ X...:/
,
.318
.3
I'
.2
EAVe../'
Ep
,
.I
I.....
o
o
40-
80" 90" 100-
" - PHASE ANGLE OF
FIGURE 9.2
120-
140"
-- 160-
180-
TRIGGERIN8-DE8REES~
HALF.WAVE PHASE CONTROL ANALYSIS CHART
233
SCR MANUAL
Oddly enough, the RMS voltage is .353 X 340 volts or 120 volts. Average voltage is .159 X 340 volts or 54 volts, which would be indicated
by a DC voltmeter across the load. Since the load resistance is 12 ohms
(1202 11200), peak current is 340/12
28.3 amperes; RMS current is
120112
10 amperes; and average current is 54112
4.5 amperes,
which would be indicated by a DC ammeter in series with the load.
Power is E Rl1S X lUllS (since the load is pure resistance) which is 1200
watts. Note carefully that E AYG X I AYG is 243 but is not true power
in the load. The SCR must be rated for 4.5 amperes (average) at a conduction angle (180 - IX) of 90 degrees. Furthermore, the load must be
able to accept the high peak voltage and current, and the line "powerfactor" is 0.5 (if defined as P LOAD +- Er,INE X ILINE RMS)'
The other alternative is a symmetrical full-wave circuit, such as
Figure 9.1(c), for which Figure 9.3 is used. The phase-angle of triggering is found to be 113 degrees for 1f4 power. Peak voltage is .92 X
340
312 volts, only a slight reduction from the half-wave case. RMS
voltage is again .353 X 340
120 volts. Average voltage is zero, presuming symmetrical wave form. RMS current is 10 amperes and power
is 1200 watts, but peak current has been reduced to 26 amperes. To
determine rating required of the two SCR's, each can be considered as
a single half-wave circuit. From Figure 9.2, the average voltage, at
113 degrees, is .097 X 340
33 volts. Average current in each SCR
is then 33112
2.75 amperes at a conduction angle of 180 - 113
= 67 degrees.
In the case of a bridged SCR circuit, Figure 9.1(e), used for this
same load, the average current through each rectifier is 2.75 amperes
but the average current through the SCR is 5.5 amperes, corresponding
to a total conduction angle of 134 degrees.
If a triac were used, its RMS current would of course be 10
amperes with a conduction angle of 67 degrees each half cycle, for a
total conduction angle of 113 degrees. This corresponds to either two
SCR's in inverse parallel or one SCR and four diodes in a bridge, but
the triac has reduced the power components to just one.
Of particular importance to note in the analysis charts is the nonlinearity of these curves. The first and last 30 degrees of each halfcycle contribute only 6 per cent (1.5% each) of the total power in each
cycle. Consequently, a triggering range from"30° to'150° will produce
a power-control range from 3% to 97% of full power, excluding voltage drop in the semiconductors .
.Figure 9.4 shows a large variety of SCR circuits for the control
of DC and AC loads, along with the appropriate equations for voltages
and currents. This .information may be used in the selection of the best
circuit for a particular 'use, and for determining the proper ratings of
the semiconductors. Figure 9.4 is fmm reference 8, which also gives
the approach for derivation of the equations shown in the chart.
=
=
=
=
=
234
=
=
AC PHASE CONTROL
EP7""'r{FULL-WAVE RECTIFIED
/
~ EPO"l
TOTAL "CONDUCTION ANGLE"=2U8O-I&>
~6~-i~~\1'--~fIj-FULL-WAVE
ALTERNATING
CI
1.0
1 I ~p~
I'
ED
.9
1:-
,
\
P:AX (R LOAD)
.8
_\
\
1\
1'1.
.7ar~
.7
\
,...
.636 -I--.6
r-..
I\.
t-...
Q:
0
~....
1\
.5
,
\
"
I'
_ERMa
Ep
I'
I"
.4
r- FULL-WAVE RECTIFIED ~AV8/'
Ep·
\
"-
.
~
.3
\
I""
.2
,',
J
"
\
o
0';;1=.
o
40"
CI
60"
80" 90" 100"
120"
140"
-PHASE ANGLE OF TRIGGERING-DEGREES
160"
18G"
FIGURE 9.3 SYMMETRICAL FULL-WAVE PHASE CONTROL ANALYSIS CHART
235
SCR MANUAL
(el
(01
(bl
NAME
CON'lECTIONS
LOAD VOLTAGE
WAVEFORMS
~~
(I) HALF-WAVE
RESISTIVE
LOAD
(2) HALF-WAVE
INDUCTIVE
LOAD WITH
FREE-WHEELING
RECTIFIER
(3) CENTERTAP
WITH RESISTIVE
LOAo,OR INDUCT·
IVE LOAD WITH
FREE·WHEELING
RECTIFIER
::JJ~
~J
~
LOAD
PEAK
FORWARD
VOLTAGE
ON SCR
PEAK
REVERSE
VOLTAGE
O~
~a~
4:
~
Jar-
MAX. LOAD
VOLTAGE
=01
(a
ED= AVERAGE
DoC IIIUJE
(el
ON
SCR
(f!
ON
DIODE
E
-
~--,-
.~
E
(~)
(d)
CIRCUIT
Ea~_~~E
EO=~
E
Ea
E
E
E
(POSSIBLY
2E If.
LOAD OPEN)
2E
E
E
=~
E =!.
o ..
E o =~
..
CRI
~
(4) CENTERTAP
WITH RESISTIVE
OR INDUCTIVE
LOAD-SCR IN
D-C CIRCUIT
(51 CENTERTAP
WITH INDUCtiVE
LOAD (NO
FREE-WtEELING
RECTiFIERI
(61SINGLE-PHASE
BRIDGE WITH
2 SCR'S WITH
COMMON ANODE
OR CATIHODE.
RESISTIVE LOAD,
OR INDUCTIVE
LOAD WITH
FREE-WHEELING
RECTIFIER
~
CR2
t
~~
JC/
LOAD
E
I
CRI
~JS
.
'E
CRI
LOAD
CRI
L
.1\7+ -~
CR2
o ~0
LOAD
R
.~
~a~
~
2E
ONCRI
E
0
-
EO=
2E
2E
-
E 0--ll.
..
E
E
E
(CRI
AND
CR21
EO= 2;
t ASSUMES ZERO FORWARD DROP IN SEMICONDUCTORS Wt£N CONDUCTING,AND
ZERO CURRENT WHEN BLOCKING; ALSO ZERO a-c LINE AND SOURCE REACTANCE.
INDUCTIVE d-c LOADS HAVE PURE d-c CURRENT.
FIGURE 9.4 CIRCUIT CONSTANTS OF SOME MAJOR PHASE CONTROLLED CIRCUITS FOR
DC LOADSt
236
2E
-;r
EON
CR2
AC PHASE CONTROL
(h)
LOAD VOLTAGE
VS
TRIGGER DELAY
ANGLE a
EO=
2~
Ii)
TRIGGER
ANGLE
RANGE
FULL ON
TO
FULL
OFF
MAX
STEADY·STATE
CURRENT
IN SCR
(I)
(k)
AVERAGE COND
AMP
ANGLE
MAX
STEADY-STATE
CURRENT
IN DIODE RECTFIER
(p)
ABILITY TO FUNDAMENTAL
PUMPBACK FREQUENCY
INDUCTIVE
OF
LOAD
LOAD
COND
VOLTAGE
ENERGY
ANGLE
TO
FOR
(I" SUPPLY
SUPPLY
MAX
FREQUENCY)
LINE
CURRENT
(nl
(m)
AVERAGE
AMP
(,)
(0)
NOTES ANO
COMMENTS
(I+C05 a)
180"
E
I
1/2
Eo= 2.,,/W(-II'-a.+2"SIN2a)
E
vR
-
180"
-
-
1
-L
EO=
2~
(I+C05«)
180"
2vR
(LOAD
HIGHLY
180"
o.S4(:R)
210·
NO
1
180"
0.26(!~ )
148"
NO
21
INDUCTIVE)
ED::
~
(I+C05a)
180"
E
-;;R"
eRr
ED:: .; (I+C05a)
180·
2E
--;;:R
= ....f...
R
CR2"o.26(!~ )
~ cos
a.
180"
E
;;R
BY RECOVERY
360"
NO
WITH HIGHLY
INDUCTIVE
LOAD
EO=
CR2 NECESSARY
WHEN LOAD IS NOT
PURELY RESISTIVE.
FREQUENCY LIMfFED
100·
180"
-
21
OF RECTIFIERS
148"
-
CHARACTERISTICS
AND SCR.
YES
21
(ASSUMING CONTINUOUS CURRENT
IN LOAD)
E
CRI=.. R
ED=~ U+COSa)
100·
E
.. R
WITHOUT CR2. SCR's
MAY 8E UNABLE
TO TURN OFF AN
INDUCTIVE LOAD.
180"
180·
NO
CR2"o.26(~)
.. R
148"
21
ALSO. CR2 R£l.£NES
SCR's FROM
FREEWHEELING
DUTY•
FIGURE 9.4 (CONTINUED)
237
SCR" MANUAL
Idl
101
CIRCuIT
101
PEAK
LOAD VOLTAGE
Ibl
FORWARD
VOLTAGE
ON seR
WAVEFORMS
CO\II\£CTIONS
NAME
(7) SNGLEMPHASE
BRIDGE WITH 2
SCR'S ON
COMMONA-C
LINE. RESiSTIVE
OR NDUCTIVE
LOAD
(B) SINGLE-PHASE
BRIDGE WITH 4,·
SCR'S AND
INDUCTIVE LOAD
" 60·
(60· < IX < 120·)
FIGURE 9.4 (CONTINUED)
239
~
(,)
(.)
(d)
QRCUtT
(a)
PEAK
LOAD VOLTAGE
WMFORMS
(bl
VOLTAGE
'"' SCR
REVERSE
VOLTAGE
EIfIflERAGE
(e)
ON
(f)
ON
SCR DIODE
~ oM
jat
(14ITHREE-.....
BRIDGE WITH
, SCR's WITH
INDUCTIVE
LOAD
LOAD
L
•
(10) TRIAC
I-
.roE
-
o-c IOWI£
Ea" RMS
Ire....,E
S./!E
ED"-w-
LOAD VOLTAGE
va
TRIGGER DELAY
ANGLE II
ED" 5":£ COS«
c..
(J61SCR INSI)E
illiDGE WIllI
A-C R!9ISTIVE
STEADY~STATE
MA'
STEADY·STATE
CURRENT
IN SeR
CURRENT
IN DIODE RECTIFIER
(11)
RANIlE
fill,",
(')
TO
FUU. AVERAGE
OFF
.IIP
120·
.fiE
(ASSUMING CONTINUOUS
CURRENT ,. LOAD)
RESISTANC£)
-;;;-
(I)
=.
120'
AVERAGE
AMP
-
(,I
COM>.
ANGLE
FOR
~
-
ABlLITYTO
"""'""'"
INlUCTlVE
LOAD
ENERGY
TO
-
(,)
en
NOTES AND
COMMENTS
:z
(,)
C"J
:::11:1
FREQUENCY
OF LOAD
VOLTAGE
(f =SUPPLY
~)
SUPPLY
LINE
Y£&
If
SCR'S REQUIRE
TWO GATE SIGNALS
W AMoRT EACH
CYCU:, ALTEftNATEC
A GATE SIGNAL
DURAT~>'o-
R
~~
+[@ ~
O§~I
E
E
-
::!lal-
~
R
,
. E "...L
ar-
Ea" ~( .. -a+
t 8ff. 2a)"2
nR
ISO'
-
-
-
f
AS WELL AS R
wR
E
0
E
E.l
a if!
E.=Ji; (,..-ca +i'S'N 2a)1I2
UR
·OR
.n.
wR
SCI'
--
FIIURE 9.4 (CONTINUED)
AND II.
Eo
180'
WITH tIDUCT1VE
LOAD. LOAD
ANO QJRRENT
DEPEND ON to! LlR
va..,.
Ea
110'
OR.!..
LOAD
-
....
"'ANGLE
(p)
MA.
(Il
A
MRAU..!LscRs
WITH RESISTIVI
LOAD
LOAD
.fiE
(I.5E IF
_&
SHJNTED
BY
(hi
VOLT...
(a "0)
FORWARD
COttWECTIONS
NAME
MAX LOAD
PEAK
Ea
ISO'
m
OR
L
"R
ISO'
-
f
INWCTANCE IN D-C
QRCUIT MUST BE
MlNMJM. FREUNCY
LIMtT DETERMINED
BY RECOVERY
DfARACTEfOIS11CS OF
REClRJIS~~
WlTHINflIJCTI\ot:
I..OAD VDLTAGE AND
_DEPEND
If L/R AS WELL
ASRAN:la.
ON
§5
iE
AC PHASE CONTROL
9.2.1 AC Inductive Load Phase Control
The above discussion considered a phase controlled resistive load
and using Figures 9.2 and 9.3 derived the required infonnation to
properly size the SCR's. In the real world most loads have some
amount of inductance. Motors, solenoids; transfonners and even some
"resistive" heaters have inductive components as a part of their impedance. The effect of this reactance is that the RMS to average current
ratio is lowered. In lowering this ratio, the dissipation of the device is
also lowered and higher average currents can be safely passed through
the SCR.
Figure 9.5 shows inverse parallel SCR's controlling a resistive
load. Also shown are the associated current and voltage waveforms.
The average current through either SCR is the average of that portion
of the load-current waveform either above or below zero. It is on this
wavefonn of current that the SCR rating is based.
SOURCE
VOLTAGE Of--~r----I---+--~--
--t----.. .-.. .---
VOLTAGE
ACROSS O f - -....
SCR'S
VOLTAGE
ACROSS 0
LOAD
~.............- _ - , - - . - - -...... , . . . - - -
LOAD O~......~-_...,.--...--........, . . . - - CURRENT
FIGURE 9.5
FULL·WAVE PHASE CONTROL OF RESISTIVE i.OAD
Figure 9.6 shows such a rating. This is a curve of the average
current .versus maximum case temperature for a 235 ampere (RMS),
C180 type SCR. Note that for different retard angles (nonconducting
portion of the cycle) the maximum-allowable average current differs.
This is due to the form factor (RMS/avg) changing with retard angle.
Figure 9.7 shows how fonn factor changes with retard angle for a
resistive load. Let's coordinate Figures 9.6 and 9.7. Note from Figure
9.7 that at 0 degree retard angle (full-cycle conduction) the form factor
is 1.57. In order to maintain the maximum rated 235 Amps (RMS), the
241
SCR MANUAL
average current must be limited to 150 amperes average (235/1.57 =
-150). At 150 degree retard angle, the form factor is approximately 4,
thus dictating a maximum average current of approximately 60 Amps.
,
V
/
-
r-I
~
~
~
~
~
w
~
m
~
~
~
0
20
40
---•
80
J
/""
100
120
140
160
RETARD ANGLE (DEGREES)
AVERAGE FORWARD CURRENT (AMPS)
FIGURE •• 6 AVERAGE FORWARD CURRENT
VERSUS CASE TEMPERATURE FOR
"PE C180 SCR
FIGURE 9.7 FORM FACTOR VERSUS RETARD··
ANGLE FOR PHASE·CONTROLLED
RESISTIVE LOAD
If the load is slightly inductive, as in Figure 9.8, the waveforms
change as shown. Note that the current waveform has been "softened"
considerably. As expected, this softening improves (lowers) the form
factor because the peak of the current waveform is reduced and its
duration extended. Figure 9.9 shows the variation of form factor with
5
-.l RETARD
I.ANGLE
R
4~-r--r--+--+-+-~-~
L
It:
~
~ 3~-~-~-t.~~~Y--~
:IE
It:
...o
VOLTAGE
ACROSS 0 I-"-"""'~"",,~""""~....,r+_..o.:.
LOAD
~
~
~O
IW
MO
I~
~,
RETARD ANGLE (DEGREES)
FIGURE 9.. FORM FACTOR VERSUS RETARD
ANGLE FOR PHASE-CONTROLLED
LOADS OF DIFFERENT POWER
FACTORS
242
AC PHASE CONTROl
retard angle for loads of different lagging power factor. Note the significant improvement in form factor at large retard angles with a
slightly inductive load. At a retard angle of 150 degrees, a 25-percent
reduction (improvement) in form factor is realized by changing the load
power factor from unity to 0.9 inductive; better than 15-percent reduction in form factor is realized from a 0.98 power-factor load. Form
factor decreases even more for power factors less than 0.9. This
improvement in average-current capability at large retard angles can
be quite significant. This is particularly true when using high-current
SCR's, where 10, 20 or 40 Amps of additional capability can be a substantial economic factor. With the introduction of high-voltage (100Ov
to 2600v) SCR's, this additional current capability represents a substantial amount of additional kva handling capability.
Now you might say, "This is all fine, but I'm still saddled with the
same old rating curves." This is true, but there's a suggestion on how
to use these curves when the load is somewhat inductive. The procedure
for determining the approximate amount of increase is as follows:
A. Average Current Versus Case Temperature
1. Locate curve on spec sheet for retard angle in question.
2. Determine new maximum average current from relationship.
Iavg(max) =
where:
Irms(max) = Maximum-rms-current rating of SCR (from spec
sheet)
F pF,4
Form factor at given lagging power factor (PF)
and retard angle (~)
3, Draw maximum-average-current cutoff line as shown in
Figure 9.10
4. Plot remainder of curve by determining distance X:
=
X(°C)
= (~PF'4)
(Tc(max) - Tc(ss»
1.0, ..
F
PF,4 = inForm
factor for power factor and retard angle
question
FLO, .. = Form factor for unity power factor and retard
angle in question
Tc(max) = Max allowable case temperature
Tc(s.)
Case temperature curve from spec sheet
B. Average Current Versus Average
Power Dissipation
1. Locate curve from spec sheet for retard angle in question.
2. Mark maximum average current on curve as previously
calculated in A.2. See Figure 9.11.
3, Plot curve by determining distance Y:
=
Y (watts)
= (~PF'4) p ••
1.0,4
where:
P ss
= Power dissipation curve from spec sheet
243
SCR MANUAL
NEW MAXIMUM
AVERAGE CURREN'"
MAXIMUM
CUTOFF
CASE
TEMPERATURE
. --~-----r--X
NEW
MAXIMUM
AVERAGE
CURRENT
CUTOFF
CURVE FROM
SPEC SHEET
AVERAGE FORWARD CURRENT (AMPS) _
AVERAGE FORWARD CURRENT (AMPS)-
FIGURE 9.11
FISURE 9.10
Figure 9.12 shows a set of actual curves for the C180 used With
a 0.9 P.F. load at 150 0 conduction. Note the 25% improvement in
current carrying capability and reduced power dissipation.
MAXIMUM CASE TEMPERATURE
II
125
NEW CASE TEMP.
CU1VE
.....,;:--.....,
~ ~ ........
150 DEG.
RETARD ANGLE
..+- 77 AMPS
FROM CI80
SPEC. SHEET..
60
20
30
40
50
AVERAGE FORWARD CURRENT (AMPS)
10
70
80
FIGURE 9.12(a)
140
/
. 150 DEG RETARD/
ANGLE
120
/
FROM CI80
SPEC. SHEET-
/
~
V
~10
V1
20
-/
/
30
/ /
/
/ot-NEW POWER
DISSIPATION CURVE
40
50
60
AVERAGE FORWARD CURRENT (AMPS)
FIGURE U2(b)
244
--
70
80
77 AMPS
AC PHASE CONTROL
9.2.2 Using Thyristors on Incandescent Lamp Loads
When incandescent lamps are switched on, there is a large current surge for the first several cycles. The ratio of surge or inrush to
operating currents are, theoretically, inversely proportional to the filament's hot to cold resistance. Typical supply circuits of 100 to 200 kva
capacity have given inrush currents of up to 25 times operating current.
Time constants of inrush current decay for typical large lamps are on
the order of 5 to 20 cycles of 60 Hertz line frequency.
Figure 9.13 shows a typical oscillogram of a GE 1500 watt projection lamp inrush current.
(a) Decay of Inrush Current
(Scale: 40 A/Division)
(b) Top: Inrush Current
Bottom: Voltage Showing Switch
Closure at Approximately 85
Electrical Degrees
(Current Scale: 100 A/Division)
FIGURE 9.13
INCANDESCENT LAMP INRUSH CURRENT
From Figure 9.13 it can readily be seen that the inrush surge is an
important consideration. Thyristors typically have a capability for
inrush for ratios of inrush to operating current anywhere from 8: 1 to
12: 1. It is important to note that lamp inrushes can be considerably
higher. Table 9.1 presents lamp theoretical inrush currents for several
of the more common lamps. In order to allow the designer to properly
allow for this inrush Table 9.2 shows a list of recommendations of
which thyristors should be used for a given lamp load.
Wattage
Rated
Voltage
6
25
60
100
100 (proj)
200
300
120
120
120
120
120
120
120
120
120
115
500
1000
1000 (proj)
TABLE 9.1
Type
Vacuum
Vacuum
Gas Filled
Gas Filled
Gas Filled
Gas Filled
Gas Filled
Gas Filled
Gas Filled
Gas Filled
Amps.
Steady State
Rated Voltage
0.050
0.21
0.50
0.83
0.87
1.67
2.50
4.17
8.3
8.7
Hot/Cold
Resistance
Ratio
12.4
13.5
13.9
14.3
,",,15.5
16.0
15.8
16.4
16.9
,",,18.0
Theoretical
peak Inrush
(170 V pk)
(Amps)
0.88
4.05
9.70
17.3
19.4
40.5
55.0
97.0
198.0
221.0
Inrush Characteristics of Several Common Lamps
245
stR MANUAl
This table has been subdivided into four columns showing both
120 and 240 volt wattages. The table is then further divided into
house wiring and industrial!commercial wiring. The house wiring
column includes the limiting effect of a standard 20 ampere household
circuit, whereas the industrial!commercial column assumes a zero
impedance line. The final assumption of the table is that the lamps are
normal incandescent lights and not high brightness types such as are
used in projection systems. To use the projection bulbs the thyristor
capability must be -reduced at least an additional 10%.
PERMISSIBLE LAMP WATTAGE
120 Volt Line
Devices
House
Wiring
240 VoH Lin.e
Industrial!
Commercial
Wiring
House
Wiring
Industriall
Commercial
Wiring
Two C103's·
60
60
Two C106's·
1SO
ISO
300
300
SC35/36
360
ISO
720
360
600
250
1,200
500
600
480
1,200
960
1,000
4SO
2,000
900
1,000
600
2,000
1,200
1,200
600
2.400
1,200
2,000
1,300
4,000
2,600
SC4O/41, S1:240/241
SC141
SC45/46,SC245/246
SC146
( SC50/si,:SC2501251
SCSO
Two C45/46's
3,400
7,SOO
Two CSO/52's
·5,000
10,000
Two C350's
7,500
15,000
Two Cl7S's
12,500
25,000
Two C180's
17,500
35,000
Two C290/1's
27,500
55,000
50.000
. 100.000
, Two C530's
* Ballast Resistor MustSe Used In Series With Load-See Application Note 200.53.
TABLE 9.2
Maximum Lamp Wattage Far Thyriston
9.3 COMMUTATION IN AC CIRCUITS
Commutation of the thyristor in AC circuits is usually no problem
because of the normal periodic reversal of supply voltage. There are
cases, ho)'Vever,which can lead to failure to commutate properly as
the result of insufficient time for turn-off, or of excessive dv/dt of
reapplied forward voltage, or both. Supply frequency and voltage, and
inductance in load or supply, are determining factors.
246
AC PHASE CONTROL
Consider the inverse-parallel SCR circuit of Figure 9.14, with an
inductive load. At the time that current reaches zero so that the conducting SCR can commutate (point A), a certain supply voltage exists
which must then appear as a forward bias across the other SCR. The
rate-of-change of this voltage is dependent on inductance and capacitance in the load circuit, as well as -on reverse recovery characteristics
of the SCR's. In certain cases, an L di/dt transient may be observed
as the result of the SCR turning off when current drops below holding
current, I H • The addition of a series RC circuit in parallel with the
SCR's, or with the load, can reduce the dv/dt to acceptable limits.
The magnitude of C is determined by the load impedance and the
dvI dt limitation of the SCR. The value of R should be such as to damp
any LC oscillation, with a minimum value determined by the repetitive
peak SCR current produced when the SCR's discharge the capacitor.
Design data for dvI dt protection circuits is covered in depth in Section 16. 3
RI
Ilel
r"V\,l'l,-1t"1
I
I
~
l
~
FIGURE 9.14
~/.
SUPPLY~VOLTAGE
:"}
LOAD
R
VOLTAGE
ACROSS
SCR
SUPPRESSION OF OV/OT ANO TRANSIENTS FOR INDUCTIVE LOAO
An alternate solution is, obviously, the use of SCR's capable of
turning off in a short time with a high applied dv/dt and a high voltage. In high-power circuits, this is often the best approach because of
the size and cost of adequate RC networks.
Inductive AC loads in the bridged SCR circuit of Figure 9.15
have a slightly different effect. The rapid reversal of voltage at the
input terminals of the bridge rectifier not only represents a high dv/dt,
but it also reduces the time available for commutation. If the rectifiers
used in the bridge have slow reverse-recovery time, compared with
turn-off time of the SCR, the reverse-recovery current is usually enough
to provide adequate time for commutation. Where this is not practical,
a series RIC I circuit at the input terminals of the bridge may be used.
An atIernate form of'suppression is to use R2 C 2 across the SCR, which
will limit dv/dt, but a resistor Ra is then required to provide a circulating current path (for current on the order of I H ) to allow sufficient
commutating time. If capacitor C 2 is large, it can provide holding current to the SCR during the normal commutation period, and thus
prevent turn-off until the capacitor is discharged.
247
SCR MANUAL
...
R3
r-·-~--------r-----
I
LOAD
: "LIi'
------~
1
~R2
II
L
VOLT,"!l.
ACRO;S
;;kC2
~dvldt
•t
-II-TURN-OFF TIME
I
I
' - -_ _ _- ' _____ JI
FIGURE 9.15 SUPPRESSION OF DV/DT AND INCREASING TURN·OFF TIME
The inductive AC load has a similar ,effect upon the commutation
of the triac, and the solution is to either obtain a faster triac or suppress
dv/dt with an RC network. This is discussed further in Section16.3·
Inductive DC loads often require the addition of a free-wheeliDg
diode, Dl in Figure 9.16, to maintain current How when the SCR is
OFF. When an inductive DC load is used in the bridge circuit, Figure
9.16(c), the inductance causes a holding current to How through the
SCR and the bridge rectifier during the time line voltage goes through
zero, preventing commutation. The addition of a free-wheeling diode,
D!> is required to by-pass this current around the SCR. The average
current rating required for the diode Dl is lh maximum average load
current for (a) and 1f4 maximum average load current for (b) and (c).
SCR
lal
leI
FIGURE 9.16
248
FREE·WHEELING DIODE BY·PASSES HOLDING CURRENT
AC PHASE CONTROL
9.4 BASIC TRIGGER CIRCUITS FOR PHASE CONTROL
Any of the relaxation oscillator pulse generators described in
Chapter 4 may be adapted to phase control work. Since these are simply
timing circuits, provision must be made for synchronizing them with
the AC supply. This is usually done by taking the oscillator input voltage from the supply. There are many ways of connecting the various
versions of the basic oscillator circuit, using the different semiconductor
triggering devices and the thyristor, supply, and load ·circuits. Each
combination has unique properties which must be considered in the
selection of a circuit to perform a desired function.
9.4.1 Half-Wave Phase Control'
The circuit of Figure 9.17 uses the basic relaxation oscillator to
trigger the SCR at controlled triggering angles, i%t, during the positive
half-cycles of line voltage. Since the adjustable resistor Rl may go to
zero resistance, diode Dl is used to protect the triggering device and
the gate of the SCR during the negative half-cycles. Certain triggering
devices will permit the use of a fixed resistor R2 instead of the diode,
as will be shown later.
The waveforms of supply voltage, e, and voltage Ve , across the
capacitor are shown in Figure 9.18. The magnitudes of R l , C, Ep, and
V s determine the rate of charging the capacitor and the phase angle, OCt,
at which triggering occurs. The earliest and latest possible triggering
angles which can be obtained are indicated by OCl and OC2 on the waveforms of Figure 9.18. If the switching current, Is (see Chapter 4), of
the trigger device is considered, the following relationships exist:
Vs = Ep sin OCl
(9.1)
and
(9.2)
Since the maximum useful value of Rl produces triggering at OC2,
Rl may be calculated for given values of e, C and V s, but ignoring Is
for the moment, using the following equation:
(9.3)
Conversely, the peak voltage across the capacitor is
V CP
V
=~
wR C +
l
0
(9.4)
249
SCR MANUAL
0'
sus tetc.}
•
~
sC •
Epsinwt
At;
SUPPL.Y
D2[ T
,,
Vc
.3
f
FIGURE 9.17 BASIC HALF·WAVE PHASE CONTROL CIRCUIT
,. ___;E~,_____ ':___ _
~
VomIlHt"GK___
-------
--
II t !l2
III
I-- SCR OFF ---I ~c: I-
",
'~;
(.J
FIGURE 9.18 WAVE FORMS FOR FIGURE 9.17 (WITH D, AND SUS)
Equations 9.3 and 9.4 assume a low value of Vs compared with
E p , as would be the case when using an SUS trigger device on a 120
volt Ae line.
From equation 9.4 it can be seen that the residual (or initial) voltage, V0, left on the capacitor has a pronounced effect upon this simple
trigger circuit. The residual voltage, V0, is usually the sum of the minimum holding voltage, VH, of the trigger, and the gate-to-cathode source
voltage, VGK, which appears when the SeR turns on.
If the switching voltage is not reached during one positive halfcycle, the trigger device does not switch and a high residual voltage is
left on the capacitor. The result, as shown in Figure 9.18(b) is "cycleskipping" as the capacitor continues to charge each positive half-cycle
until the trigger device switches. If the range of Rl can be limited so
that triggering always occurs each half-cycle under worst-case tolerance
conditions of minimum E p , minimum e, maximum V s, minimum Is,
and minimum V0, this cycle-skipping can be avoided. On the other
hand, with the opposite tolerance conditions, the latest possible triggering angle may produce an unacceptable minimum power in the load.
The ultimate solution to cycle-skipping is to automatically reset
the capacitor to a known voltage, V0, a the end of each half-cycle even
though Vs was never reached. One way of doing ,this is to substitute
resistor R2 in the place of diode D2 in Figure 9.17. This causes the
capacitor voltage to reverse on the negative half-cycle, yielding a nega-
250
AC PHASE CONTROL
tive value of V0 at the start of the positive half-cycle. If the triggering
device does not conduct during the negative excursion of Vc, then V0
will be predictable for any given value of R I . This connection provides
one cycle for the residual voltage on C to decay, and eliminates cycle
skipping. If the triggering device conducts when Vc is negative, a
second diode, D 2 , may be used to clamp Vc to approximately -1 volt
during the negative half-cycle.
If a bilateral trigger is used, such as the SBS or a Diac, the
diode D2 is not required (provided R2 adequately limits the negative
current) but V0, at the beginning of the positive half-cycle, will depend
on the number of oscillations occurring during the negative half-cycle,
hence upon setting of R I . Changing RI will make an integral change
in the number of negative oscillations, hence will make step changes
in Vo. This action results in step changes in triggering angle.
Automatic reset of capacitor voltage is achieved in the circuits of
Figure 9.19 by forcing the triggering device to switch at the end of the
positive half-cycle. In circuit (a), resistor R2 provides a negative current
out of the gate of the SUS (see Chapter 4) when the line voltage goes
negative, thus causing the SUS to switch and discharge the capacitor.
LOAD
440W
f---..,--.,-----,
(e80W)
01
GE
AI48
120 VAC
(AI4D)
120 v
(240V)
60Hz
A2
2201<
(470K)
AI
(240 VAe)
SOOK
(I MEG)
1
Ib)
CI
Q.2JAof
$CR'GE C20B
(C20D)
01' GE 2N2646
la)
RO'
33~6~~~S
RI'50K OHMS
e,o.1 MFD
RBI'470HMS
DI=AI4B(AI4DI
NOTE: VALUES IN PARENTHESES APPLY FOR 240 VAC SUPPLY
FIGURE 9.19
HALF-WAVE PHASE CONTROL WITH CAPACITOR RESET
Since the switching voltage of the unijunction transistor is a function (TJ) of the interbase voltage, the capacitor in circuit (b) is reset
through the UJT at the end of the positive half-cycle when the interbase voltage dips toward zero.
In the preceding examples, the supply voltage for the triggering
circuit collapses when the SCR turns on. This connection avoids multiple oscillations and permits decreasing R} to zero without damage to
the control circuit. If the triggering circuit were connected directly to
the supply voltage instead of to the SCR anode, a fixed resistor, approximately 5000 ohms, in series with RI would be required to limit current.
Since this latter connection changes the control circuit· from a twoterminal to a three-terminal circuit. wiring considerations in certain
applications may prohibit its use.
251
SCR MANUAL
8.4.2 Full.Wave Phase Control
Either of the half-wave control circuits of Figure 9.19 may be,
used for full-wave power control by connecting them "inside" the
bridge rectifier circuit shown in Figure 9.1(e). The UJT circuit of Figure 9.19(b) requires no modification for this use, but the SUS circuit (a)
requires changing R2 to 22 K ohms, adding another 22 K ohms between
gate of SUS and cathode of SCR, and deleting diode D 1 • These revisions are needed to obtain the reset action of the SUS.
The most elementary form of full-wave phase control,js the simple
diac/triac circuit of Figure 9.20. The waveform- of capacitor voltage,
Vc in Figure 9.21, is quite similar to the half-wave case with the major
exception that the residual capacitor voltage, V0, at the start of each
half-cycle is opposite in polarity to the next succeeding switching voltage, Vs, that must be reached. The waveform shown for Vc is a steadystate condition, triggering late in each half-cycle. If the resistor Rl is
increased slightly, the dotted waveform, Vc/, shows what happens in
the next cycle after the last triggering. At the start of this cycle, V0 is
the same as steady-state since the diac had switched in the preceding
half-cycle. At the end of the first half-cycle, however, the capacitor
voltage is just below Vs, and the diac remains dormant. This changes
Vo to +Vs at the beginning of the second half-cycle. The peak capacitor voltage in the negative half-cycle is, therefore, consider~bly below
Vs, as shown earlier by equation 9.4. In all succeeding cycles V0
VCp
and the peak value of Vc will then remain well below Vs until the value
of Rl is reduced.
=
72rJW
(1440W)
01
250K
ISOOK)
I20Y
(MOV)
60",
T
c
Y
1
DIAC
GE ST-2
(-!vo -
CI
O. I,.. f
NOTE:
VALUES IN PARENTHESES ARE FOR 240VAC SUPPLY.
FIGURE 9.20 BASIC DIAC·TRIAC FULL·WAVE
PHASE CONTRO,l
FIGURE 9.21
WAVEFDRMS FOR FIGURE 9.20
Once triggering has ceased, reducingR1 will raise Vc, but when'
Vs is reached again and the diac switches,· V0 is suddenly reduced.
This action increases the value of Vc on the next half-cycle, which·
causes triggering to occur at a much earlier phase angle. As a result,
the load current suddenly snaps from zero to some intermediate value,
from which point it may be smoothly controlled' over the full range
from al to a2'
252
AC PHASE CONTROL
The "snap-on" effect may be eliminated by using the ST4 asymmetrical silicon bilateral switch (ASBS), as shown in Figure 9.22. It was
shown that the snap-on of the diac-triac phase control of Figure 9.20
was due to the fact that the capacitor was charging through a voltage
of two times Vs each half cycle, but when the diac triggered the offset
caused the capacitor to reach Vs earlier in the cycle. The ASBS has
been designed to use this offset to an advantage. Figure 9.23 shows
how this is accomplished. (Remember that the ASBS breakover voltage
is about 8 volts in one direction and twice that in the other direction.)
It can be seen that if Rl of Figure 9.22 is set so that the ASBS can trigger at point A, the capacitor is essentially uncharged at the zero voltage crossing following point A. If the ASBS were symmetrical, it would
indeed switch earlier in the next half cycle (at point C). But since the
breakover voltage in that direction is twice that at point A, the capacitor continues to charge until point B. At this point the ASBS triggers
and delivers half the capacitor's charge to the triac gate. At this time
the capacitor is at the same voltage it was at before the ASBS triggered
at all. The result is that the snap-on has been reduced to an almost
negligible value with no increase in component count. Since some waveform asymmetry is present in the ASBS phase control circuit, its use
may not be practical to drive loads where no significant de component
can be tolerated, e.g. fluorescent lamps, transformers, primaries, and
the like.
MAXIMUM
LAMP
LOAD
IOOOW
(2000w1
TRIAC
120V
(240VI
60Hz
t-ffi:l+--><..i/
GE SCI468
(5CI460)
NOTE:
VALUES IN PARENTHESES ARE
FOR 240VAC SUPPLY.
FIGURE 9.22 FULL·WAVE ASBS·TRIAC
CONTROL FOR NEGLIGIBLE SNAP·ON
FIGURE 9.23
WAVEFORMS OF ST4 AT SNAp·ON
Figure 9.24 shows another circuit with very little snap-on effect.
This circuit uses a second capacitor, C 2 , to recharge C1 after triggering,
thus raising V0 to approximately V s. The maximum, or latest, triggering angle, (%2, with this circuit is not limited to the point where the
supply voltage' is equ!ll to Vs because the second capacitor will permit
greater than 90 0 phase shift of VCl' If, however, the diac should switch
after the 180 0 point on the supply wave, it could very well trigger the
triac at the beginning of the next half-cycle. Since this condition usually
needs to be avoided, coupling resistor Ra should be adjustable to. permit
compensation for wide-tolerance component values. If desired, Rs may
be set for a minimum power level in the load at maximum setting of RIo
253
SCR MANUAL
¥' ALTERNATE
1200W
(2400W)
LOAD POSITION
~~~r_-_-_'+--1-----------'-------i
:
.2
~
68K
HSOKJ
fZOV
(240V)
50 OR 60Hz
j
1
TRIAC
GE
5C 1458
(SC 14501
"---__----'""-r
I
'
I
I
I
:
_J_
C2
"f'
Qlp.f
-----------~-------j
dV/dt SUPPRESSOR
AS REOUIRED
FIGURE 9.24
EXTENDED RANGE FULL·WAVE PHASE CONTROL CIRCUIT
9.5 HIGHER "GAIN" TRIGGER CIRCUITS FOR PHASE CONTROL
All of the previous circuits control phase angle of triggering by a
resistor. To control over the full range, from minimum to maximum
power, with the simple RC timing circuit requires a very large change
in the value of R, presenting a low control "gain." For manual control,
this is most adequate. For systems which must perform a function, in
response to some signal,. the simple RC circuits are usually inadequate,
although a photoconductor or a thermistor could be used for control
but only over very wide range of light or temperature change.
9.5.1 Manual Control
Figure 9.25 shows a conventional, manually-controlled triac circuit with a unijunction transistor. A zener diode clamps the control
circuit voltage to a fixed level, as shown in Figure 9.26. Since the peakpoint (or triggering) voltage, ell> of the unijunction transistor emitter is
a fixed fract;"!l of the interbase volage, V BB, as indicated by the dashed
curve, the capacitor will charge on an exponential curve toward V BB
until its voltage reaches e p • Assuming, for convenience, that e p is 0.63
VBB, triggering will occur at one time-constant. Therefore, to cover the
range from 0.3 to 8.0 milliseconds, the product R 2 C must change by
the same amount. Since C is fixed, R2 must then be varied over a 27: 1
range. Not only is this a very large range, but the transfer characteristic
from R2 to average load voltage, VL , is quite non-linear, as shown in
Figure 9.27. These characteristics are usually satisfactory, however, for
manual control.
Replacing the manually controlled resistor with a p-n-p transistor,
shown in Figure 9.28(a), and applying a DC signal between emitter
and base results in a higher current-gain but the range of base current
must again be 27: 1. The transfer characteristic, Figure 9.28(b), also
remains non-linear.
254
AC PHASE CONTROL
.,
1200W
[2400W)
02
03
6800
2W
U2K,5W)
04
05
01 = GE-14X20
°2,3,4.,5" GE IN5059 (GE IN5060)
FIGURE 9.25
CONVENTIONAL PHASE·CONTROL CIRCUIT
FIGURE 9.26
UNINJUNCTION TRANSISTOR WAVEFORMS
100 i'\
80
\
60
'WL
40
20
'I
\
20
FIGURE 9.27
\
i\
,
40
60
·3
80
IOOx 103 n
TRANSFER CHARACTERISTIC OF CONVENTIONAL CIRCUIT (FIGURE 9.25)
R2
QI
(0)
FIGURE 9.28
100H
00 '80
60
""1IL40
'
'
20
'
Ib)
SERIES TRANSISTOR CONTROLLED RAMP
255
SCRMANUAL
Control gain can be made very high by the use of a low resistance
potentiometer, connected as shown in Figure 9.29(a). Since the exponential charging of C is very fast, and limited by the voltage-division
of the pot, the transfer characteristic is again non-linear, as shown in
Figure 9.29(b). If the zener clamp has any significant zener impedance,
the clamped voltage will not be Hat, but will have a slight peak at 90
degrees. This curvature can produce an abrupt discontinuity, or "snap,"
in the transfer characteristic as indicated by the dashed curve of
Figure 9.29(b).
loom
80
. ,
60
:
%VL 40
:
20
'
00
204060~1OO
%R
(al
FIGURE 9.29
(b)
RESISTANCE CONTROLLED PEDESTAL
The use of an n-p-n transistor, Figure 9.30(a), will provide a high
current-gain, but non-linearity and possible snap are still present, as
indicated in Figure 9.30(b).
Ra
R2
2.2K
80 0 0 .
1
%VL:
20
00
(al
FIGURE 9.30
I
ie
(bl
SHUNT TRANSISTOR CONTROLLED PEDESTAL
9.5.2 Ramp-and-Pedestal Control
If the circuits of Figure 9.25 and 9.29(a) are combined with diode
coupling, as in Figure 9.31(a), the exponential ramp function can be
caused to start from a higher voltage pedestal, as determined by the
potentiometer. Transfer characteristic Curve 1 of Figure 9.31(b) is
obtained when R2 is set for a time-constant of 8 milliseconds. Higher
O.IL.-----''----'-L
(al
FIGURE 9.31
~t
(el
RESISTANCE CONTROLLED PEDESTAL WITH LINEAR RAMP
control gain is obtained (Curve 2) by making the R2 C 1 time-constant
about 25 milliseconds. The voltage wave-shape observed across C 1 is
256
AC PHASE CONTROL
a nearly-linear ramp sitting on a variable-height pedestal, as in Figure
9.31(c). Small changes in pedestal height produce large changes in
phase-angle of triggering. The linear relationship between height and
phase-angle results, however, in a non-linear transfer function because
of the shape of the sine-wave supply.
Both high gain and linearity are obtained by charging C 1 from
the undamped sinusoidal waveform, as in Figure 9.32(a). This adds a
cosine wave to the linear ramp to compensate for the sinusoidal supply
waveform, resulting in the linear transfer characteristics shown in Figure 9.32(b). System gain can be adjusted over a wide range by changing the magnitude of charging resistor, R 2 , as indicated in Figure
9.32(c). By selecting a ramp amplitude of one volt, for example, and
assuming a zener diode of 20 volts, then a change in potentiometer
setting of only 5 percent results in the linear, full-range change in
output.
The values shown in Figure 9.32(a) are typical for a 60 Hz circuit.
The potentiometer resistance must be low enough to charge capacitor
C 1 rapidly, in order to be able to trigger early in the cycle. This is the
limiting factor on control impedance level. The logarithmic characteristic of diodes limits the control gain that can be achieved with a
reasonably linear transfer characteristic. At a one-volt ramp amplitude,
diode non-linearity is not pronounced, but a 0.1 volt ramp voltage, the
eo
loon
""'lit.:
20
00 20406080 100
%R,
(b)
FIGURE 9.32
(a)
l!P..----=___\ _
\f5?F;;;;;V
0- a t
(c)
RESISTANCE CONTROllED PEDESTAL WITH COSINE-MODIFIED RAMP
capacitor is charged primarily by diode current, thus obliterating the
cosine-modified ramp. The sharper knee of a zener diode may be used
to obtain higher gains, at the expense of requiring a higher voltage
across the potentiometer. The third limiting factor is the peak-point
current of the unijunction transistor. This current must be supplied
entirely by R2 and should be no higher than one-tenth the charging
current on C 1 , at the end of the half-cycle, in order to avoid distortion
of the waveform. The 2N2647 unijunction transistor used in the example has a maximum peak-point current of two microamperes. For cases
where a lower peak point current is required, the General Electric
257
SCR MANUAL
D13T2 (2N6028), Programmable Unijunction Transistor is available
with a peak point current as low as 150 nanoamperes. The fourth limitation is the zener impedance of diode D 1 • This impedance must be very
low in order to keep the peak-point voltage (triggering level) constant
during the half-cycle. If this voltage changes 0.1 volt, then the ramp
voltage should be on the order of 1 volt. The temperature effects on the
unijunction transistor, and other components, must also be taken into
consideration when attempting to work at very low ramp voltages.
In Figure 9.33(a), manual control is replaced by a bridge circuit
for feedback control. Zener diode D2 has a slightly lower zener-voltage
than Dl in order to hold the top of the clamped waveform more nearly
Hat. Resistors Rl and R2 form the voltage divider which determines
pedestal height. Variation in either of these resistors can therefore provide the control function, although R2 is generally used as the variable.
Figures 9.33(b) and (c) show the use of a thermistor for temperature
regulation and a photoconductor for light control, in either open-loop
or closed-loop systems.
01
2(]V
FIGURE 9.33
leI
(bl
(al
OHMIC·TRANSDUCER PEDESTAL CONTROL
To obtain a higher input impedance, an n-p-n transistor may be
used as an emitter-follower, as shown in Figure 9.34(a). If the transistor
has a current gain of 100, the values of Rl and R2 can be increased
from 3000 ohms to 300 K ohms, thus greatly reducing power dissipation in the sensing element. This is particularly important when Rl or
R2 is a thermistor. Resistor Ra is required in the collector circuit of the
transistor in order to limit charging current available to the UJT capacitor and thus prevent premature triggering of the UJT.
+DC
(a)
Ibl
FIGURE 9.34 TRANSISTOR EMITTER·FOLLOWER CONTROL FOR AC OR DC INPUT
258
AC PHASE CONTROL
In many feedback-control systems, high gain and phase-shifts ol:ten
produce instability, ranging from excessive overshoot to large oscillations, or hunting. The transistor permits use of a DC sensing circuit
followed by an appropriate RC "notch-network" (Rl Cl> R2 C 2) to produce the required degree of damping. Since the cosine-modified ramp
results in a uniform, linear response, system gain is constant and proper
damping is much easier to obtain than in the case of the linear ramp
where gain changes with phase-angle. System gain is controlled by the
ramp charging resistor (R2 of Figure 9.32), which can be made a secondary variable through the use of a thermistor or a photoconductor.
To avoid excessive loading on the DC sensing circuit, a resistor is required in series with the base of the transistor. Upper and lower control
limits may be obtained by the use of diode clamps.
The capability of working from a DC control signal permits a softstart and soft-stop circuit, shown in Figure 9.35(a). This circuit features
individually adjustable rates of start and stop, good linearity, upper and
lower limit clamps, and manual or resistive master phase control by
means of the top clamping level. For a typical UJT peak-point of 2/3
the interbase voltage, the ramp amplitude may be set at 113 interbase
voltage and the pedestal clamped at 1/3 and 2/3 this voltage. The resulting performance characteristic is shown in Figure 9.35(b) and (c)
for this condition, with the switch turned ON at t, and OFF at tao
(0)
---~·:,:~----5
....
Vc
'I '2
'2 '2 '
(~
FIGURE 9.35
Lcl
V~
'I '2
'3 '4
t
~
SOFT START AND STOP CONTROL
Remote control from an AC signal, such as an audio-frequency
from a tape recorder or from a tachometer, or an RF carrier alone or
with audio modulation, is shown in Figure 9.36. The offset voltage
characteristic of a: high-gain system provides immunity to noise and
effectively decreases the band-width of the input resonant circuit. If
offset is not desired, but high-gain is required, the input circuit may be
biased, by the dotted resistors R4 and RI), to a voltage just below offset.
The use of a standard ratio-detector will permit control by an FM signal directly.
259
SCR MANUAL
AC SIGNAL ~
INPUT o----J
FIGURE 9.3&
, -~""""--+-"""'-+i.
C"
01
FREILUENCY-SELECTIVE AC AMPLITUDE CONTROL CIRCUIT
Alternate transistor connections are shown in Figure 9.37, providing a wide variety of performance characteristics. The emitter-follower
circuit is simplified in Figure 9.37(:.) for low-gain use. At high control
gain (low ramp voltage) the emitter current requirement is very low,
and the decrease in beta at such low currents causes excessive non. linearity. Standard common-emitter connections for n-p-n and p-n-p
transistors, Figures 9.37(b) and (c), provide lower input impedance and
higher voltage gain, but require temperature compensation in high-gain
applications. In addition, the n-p-n circuit of Figure 9.37(b) results in
a sense inversion which mayor may not be desirable. Sense inversion is
also obtained in the p-n-p emitter-follower of Figure 9.37(d). The excellent performance characteristics and low cost of the 2N2923 silicon
n-p-n transistor, however, make the choice of the n-p-n emitter-follower
circuit attractive, particularly since the temperature changes have very
little effect on operation of this circuit.
:
.
J:--J:--m t
lC
..-
....-
~
-
l
._
;>
(a)
___
(b)
FIGURE 9.37
___
--
-
VC
Ic)
__ _
(d)
ALTERNATE TRANSISTOR PEDESTAL CONTRGL CIRCUITS
An alternate form of the soft-start circuit is shown in Figure 9.38
using the clamping diode, Dh to control pedestal height on a linear
ramp. Capacitor C 1 may be several hundred microfarads and is charged
slowly through R:!. Rl continues the charging beyond emitter peak-point
voltage to completely remove the effect of C 1 and to provide a discharge
path when power is removed. The supply for this circuit must be
obtained from the line, rather than from voltage across the triac, in
order to completely charge C 1 •
RI
Rz
Ra
. FIGURE 9.38
260
R4
ALTERNATE SOFT·START CIRCUIT
AC PHASE CONTROL
9.5.3 AWide Range Line-Voltage Compensation Control
In Figure 9.39, compensation for changes in supply voltage is
obtained by R2 and C 1 which add to the zener diode voltage a DC voltage proportional to supply voltage. This is used to supply interbase
voltage for the VJT. Since pedestal height is fixed by the zener diode,
reducing supply voltage reduces interbase and peak-point voltages of
the VJT, thus causing triggering to occur earlier on the ramp. The size
of R2 is dependent on ramp amplitude, hence upon Rs. The voltage
compensation feature does not interfere with use of the pedestal height
in any other control form, such as a feedback control system. This system has been found capable of holding RMS output voltage constant
within 5% for a 50% change in supply voltage. The bottom end of
control is reached when supply voltage drops to desired output voltage.
DiS
R5
RI: SOOOl1,3W
R2:50011
+
CI
R6
DI-DS:INSOS9
D6:Z0V,IW
ZENER DIODE
R3
05
04
R4
TI
QI
R3: 3300
R4: 10K
R5:5M
R6:IK
CZ
CI: 200,..1,10.
CZ:O.I,..I
QI : GE 2N2646
TI: SPRAGUE II ll2
FIGURE 9.39
WIDE RANGE LlNE·VOLTAGE COMPENSATION CONTROL
Current feedback control can be obtained by the use of voltage
across a shunt resistor, but this requires rectification and filtering when
AC is to be controlled since no current flows prior to triggering. In addition, a lamp may be used as the shunt, with a photoconductor sensing
lamp output. Response time of the lamp and photoconductor is generally long enough to provide filtering, and the control is on the square
of current, hence will hold constant RMS value rather than average
value. A resistor-thermistor combination will also provide RMS control.
A current-transformer may be used to produce a higher output voltage
signal on AC with less power loss. If power loss in the shunt is detrimental, a magnetic-flux sensitive element such as a resistance transducer or a Hall-effect element may be used in a suitable coil and core.
In these magnetic-flux sensors, the output will be a function of average
current.
These circuits are typical of a wide variety, based on the rampand-pedestal concept for transfer from voltage, current, or impedance
level to phase-angle of triggering for SCR's. Adjustable gain, linearity,
selection of high or low input impedance, and operation from a DC
input signal are . attractive features for use in feedback or open-loop
control systems, or in special function systems.
261
SCR MANUAL
9.5.4 3 kw Pbase·C.ontrolled Voltage Regulator
Figure 9.40 is shown here hecause it exemplifies a method of regulatingthe RMS value of an unfiltered phase-controlled voltage across a
resistive load.
If the voltage across the load was fed directly hack to the control
circuit and compared to a reference, there would he an undervoltage
error when the SCR's are off and an overvoltage when the SCR's were
on. Because of this unstahle condition, this hypothetical system would
not regulate. To ohtain regulation (average or RMS regulation), one
must have a stahle feedhack signal as well as a reference. In Figure
9.40 this condition is met hy using Ll and R16, shown within the right
hand dotted box. Since the light from a lamp is fairly stahle on phase
control, this light generated hy the load voltage and proportional to it
can he used as a feedhack signal that is proportional to the RMS voltage of the load. In this example the light is coupled to the photocell,
P.C. (the PL5Bl is an integral lamp photocell comhination designed
for this and similar uses). A change in the output voltage causes a .
.change in light and therefore a change in the photocell resistance. This
change in photocell resistance unhalances the resistor hridge. The
hridge unhalance then changes the pedestal level of the ramp and
pedestal phase control through use of the differential amplifier consisting of Ql and Q2'
."
r---- - ----------
I
I
I
.
'""
~~~ ~"
II
80H,
.~I
I
I
I
I
I
I
I
"
I
I
I
I
I
I
I
I
,I
TI
HLt_=_=_=_=_=_=_=_=_=_=_=_=_t_-_-"-_...._-_--J_o---l
__ J
T3-TlJRNS"ATIO/'Lft.4KVA
RI"22KA
fl2-Z.2Ka
R3-l8ltA
IM'"JOKaMULTL-TURN TRIM POT
R!rIOK.
R6-2.2K.Q
RT-2.Z1(a
Ra--U:KA
::o--~~!
RlI-8.ZKIl,IOWATT
Rl2-1.5MEGn
RI3-IKA
m4,R15-22a2W
11I6-9K:t.5'l1o,14WATTS
RI7-33a.tW
RI8-334,2*
LI-INCANDESCINT LAMP OF SE PL581 (LA .... -PHOfO(:ELU INTEGRAL t1EfIllETlCAlLY SEALED UNIT
P.C.-CAIlIifIUM.stJl.FIlf PHOTOCELL OF' SE PL~ tLAMP-f*«)1OCEUI INTEGRAL !£RMETlCALLY SEALED t.WoItT
Ala. RESISTORS t ,,",-AND 112 "WATT EXCEPT WttERE NllCATED
FIGURE 9.40 3 KW PHASE·CONTROLLED··YOLTlGE REGULATOR
The system also has other important features which should he
mentioned. For example, hy placing a low resistance in series with the
load and in parallel with the feedhack lamp L], as shown in Figure
9.41, a current regulator is formed. In this circuit, the current through
the load is maintained at a constant value.
262
AC PHASE CONTROL
LOAD
T3
LOW
SENSING
R
FIGURE 9.41
PHOTOCELL
LAMP LI
CIRCUIT CHANGES FOR CURRENT REGULATOR FOR FIGURE 9.40
I
~
RII
2.2K
r- - - - -+- - -
RI2
I
"1
B+
I
CR6
5.6K
03
CR5
I
I
47K
...L
T2
C2
'T'
I
I
I
2.2K
FIGURE 9.42
SOn-START CIRCUITORY FOR FIGURE 9.40
An additional feature is that the 3 kw load could be a lamp. This
would eliminate the need for the feedback lamp L l . In this case the
photocell could monitor, and the system would regulate, the light output of the load. If this were done, the soft-start circuit of Figure 9.41
might be necessary.
Note also C 1 and R5 in Figure 9.40. These components constitute
a "notch" network necessary for stabilization.
Figure 9.42 shows the regulation obtained by the voltage regulator with a 3 kw resistive load for a regulated true RMS load voltage of
300 volts and a nominal input line voltage of 220 volts RMS, 50 or
60 Hz.
Input Line Voltage
True RMS
Load Voltage
Load Voltage
Change
220 V RMS (Nom)
190 V RMS
250 V RMS
300 (Nom)
Approx. 299.0
Approx. 299.0
«0.33%)
(0.33%)
FIGURE 9.43
-
<
Response Time
Less Than
100 msec. for
step change
in input
TABLE OF REGVLATION FOR CIRCUIT OF FIGURE 9.40
263
SCR MANUAl
9.5.5 860 Watt Limited-Range Low Cost Precision Light Control
The system of Figure 9.44 is designed to regulate an 860 watt
lamp load from half to full power. This is achieved by the controlledhalf-plus-fixed-half-wave phase control method. Half power applied to
an incandescent lamp results in 30% of the full light output. Consequently the circuit is designed to control the light output of the lamp
from 30% to 100% of maximum.
The operation of the closed loop is straightforward with the major
features being the load Ll and L2 , the error signal, ·photo cell, P.C., the
feedback element, Q2 the error detector, and Ra the reference.
The method of obtaining the controlled-half-plus-fixed-half-wave
is easily seen by realizing that Dl and Ql are in inverse parallel and in
series with the load. Also note that Ql will turn on during the positive
half-cycle at a time dictated by the feedback elements, reference, and
error detector located in the dotted block. Consequently, the positive
half-cycle is controllable. The function of the dotted block is identical
to the ramp and pedestal control of Figure 9.31. Now note that Dl will
conduct during the entire period of every negative half-cycle. Therefore, the negative half-cycle is continually applied to the load. This
configuration results in a controllable positive and fixed negative halfcycle applied to the load. It is interesting to note that when Dl conducts
during the negative half-cycle it resets the unijunction firing circuit.
Again temperature and voltage stability of the dotted block is
achieved by Zl, common voltage references, and the stability of the
unijunction.
This method of phase-control results in an unsymmetrical wave
with a resulting DC component. Therefore, this waveform is not suitable for transformer-fed loads.
The system will regulate the light level to within ± 1 % for a
±IO% chan2e in amplitude of the supply voltage.
186OW)
430W
R,
,----------------,
I
I
I
430W
(240)
'20Y
60Hz
1860W)
I
I
z, I
I
I
I
I
I
'- ________________ JI
R,-6.8KD.2W !l5K,5W)
R2 -47D
R3 -'KD
R4 -680D
R5 -IKA .1/2W POT
P.c.-GE A35
O,-tiE CI2281C122D)
02- GE 2N2646
D,-GE A4IBIA4'D)
~ -. 03,.fd,25V
ez-CII,.fd,25V
NOTE' ALL RES'S1lINCES
Dz-GE A'4F
Z,-GE Z4XL'6
112w,IO% UNLESS
L,.L 2 - 430W 'NCANDESCENT LAMP
FOR 240V
OTHERWISE NOTED.
VALUES 'N PARANTHESES
FIGURE 9.44 ,&OW, LIMITED RANGE, LOW·COST PRECISION LIGHT CONTROL
264
AC PHASE CONTROL
9.6 TRIGGER CIRCUIT FOR INDUCTIVE AC LOADS
Inductive AC loads present two basic requirements of the trigger
circuits in order to provide symmetry and proper control: a) synchronization must be obtained from the supply voltage rather than SCR voltage; b) the trigger signal must be continuous during most of the desired
conduction period. Figure 9.45 shows a trigger circuit specifically
designed to meet these requirements.
Unijunction transistor Ql is connected across the AC supply line
by means of the bridge rectifier, CR1 through CR4 , thus permitting Q1
to trigger on both halves of the AC cycle.
120 VOLT
AC SUPPLY
60CPS
SCR2
TRANSFORMER OR OTHER INDUCTIVE LOAD
RI-3.3K ,5 WATT
R2- 250K, 2 WATT
R3- 3.3 K, I WATT
R4-330, 1/2 WATT
R5,R6- 22A, 2 WATT
R7,RB- 33A, 2 WATT
RS,RIO- 41A, 1/2 WATT
CI,C2,C3- 0.1 MFD
QI- GE 2N2646
SCRI, SCR2 CONTROLLED RECTIFIERS, AS
REQUIRED
SCR3,SCR4- GECI06FI OR CI03Y·
TO LOAD
CRI TO CR4 - GE IN5059
CR5,CR6- GE INI165
CR1- GE AI4F
CRB- GE INI116
TI- ISOLATION TRANSFORMER 120/12.6112.6 VAC;
PRIMARY VOLTAGE DEPENDS ON LINE
VOLTAGE (UTe FT-IO FOR 120V.)
T2- PULSE TRANSFORMER PE2229, UTC H51 OR
SPRAGUE· .ftZ13 EQUIVALENT
FIGURE 9.45 TRIGGER CIRCUIT FOR PHASE CONTROLLED SCR's FEEDING INDUCTIVE LOAD
The time constant of potentiometer R2 in conjunction with capacitor C 1 determines the delay angle a at. which the unijunction transistor
delivers its first pulse to the primary of pulse transformer T 2 during each
half-cycle. These pulses are coupled directly to the gates of SCRa and
SCR4 • Whichever of these SCR's has positive anode voltage during that
specific half-cycle triggers and delivers voltage to its respective main
SCR, firing it in turn. The low voltage AC supply for the "pilot" SCR's
(SCRa and SCR 4 ) is derived from a "filament" type transformer T l'
Zener diodes CR5 and CRe, in conjunction with resistors R5 and Re, clip
the AC gate voltage to prevent excessive power dissipation in the gates
265
SCR MANUAL
of the main SCR's. The RC networks (R 7-C 2 and Rs-C a) also limit
gate dissipation in the main SCR's while delivering a momentarily
higher gate pulse at the beginning of the conduction period to accelerate the switching action in the main SCR's.
If electrical isolation of a DC control signal from the AC voltage
is required, the entire unijunction trigger circuit with its bridge rectifier and associated components can be connected to an additional secondary winding (approximately no volts) on transformer T 1 . Total
loading of this particular part of the circuit is less than 30 milliamperes.
Of course, low level phase control signals for SeRa and SCR 4 can be
secured from other circuits than the specific one shown, but this is
incidental to the main objectives: driving SCRl and SCR2 from a square
wave source synchronized to the AC line.
Figure 9.46 is a smaller and lower-cost version of the inductive
load phase control. The bridge rectifier DrD4 supplies power to the
UJT trigger circuit and supplies holding current to the SCR. If triggering should occur prior to turn-off of the triac, the SCR will be turned
on and held by current through R l • When the triac turns-off at a current
zero, triac gate current will How, depending on polarity, through Do or
D 6 , the SCR, and D4 or D 2 , thus re-triggering the triac. At the expense
of higher voltage diodes and SCR, this circuit eliminates all transformers with their attendant cost, size and weight.
RI
6800
2W
01
02
R2
6800
2W
05
120V
50 OR
60Hz
R3
10
03
R4
22
04
SCR
GE
CI06B
GE
Z4XI6
01-D6zGEIN!5059
R5
100
06
C2
0.1
NOTE' FOR 240 VOLT OPERATION'
RI-R2-15K.4W
SCR-GECI0601
01- D6. GEIN5060
R5,C2'dv/dt SUPPRESSION
FIGURE 9.46 FULL·WAVE PHASE CONTROL FOR INDUCTIVE LOADS
266
AC PHASE CONTROL
For higher power-factor inductive loads, and where a small dissymmetry is permissible, the two-capacitor diac/triac circuit of Figure
9.24 may be used, with the load connected in the alternate position
shown in that circuit. When Rl is small, calling for maximum power,
the trigger circuit supply is essentially the voltage across the triac,
hence cannot attempt to trigger before commutation. When Rl is large,
the trigger circuit is largely powered from the supply voltage, thus providing good symmetry and very little DC component of load current.
9.7 PHASE CONTROL WITH INTEGRATED CIRCUITS
Up to this point the trigger circuits described have been fairly
simple and straightforward, but it can be seen that the component
count can get rather high. To achieve high performance the component tolerancy can cause the design to become cumbersome and
expensive. To simplify design, while maintaining high performance,
General Electric has designed and .marketed a unique monolithic integrated circuit, the P A436 integrated phase control trigger circuit.
9.7.1 The PA436 Monolithic Integrated Phase-Control Trigger Circuit
The PA436 isa high-gain trigger circuit for phase control of triacs,
or SCR's. It is specifically intended for the speed control of AC induction motors, but can also be used on purely resistive loads such as
incandescent lamps. This circuit accepts a thermistor signal for temperature control of fans and blowers, or a DC tachometer signal for
feedback speed regulation. Adjustable gain, zener-regulated voltage,
ambient temperature compensation, and inductive load logic are primary attributes of this integrated trigger circuit.
The PA436 converts an analog input signal to a phase-controlled
pulse for triggering thyristors. The signal is compared with a reference
and the phase-angle of triggering is obtained by use of the ramp-andpedestal technique described earlier in this chapter.
a
TRIGGERING
ANGLE
lal POSITIVE RAMP
Ibl NEGATIVE RAMP
FIGURE 9.47
RAMP AND PEDESTAL WAVEFORM
267
8CR MANUAL
Figure 9,47(a) shows the typical ramp-and-pedestal waveform,
with positive cosine ramp, as is used in unijunction transistor· phase
control circuits. The P A436 operates with a negative cosine ramp, as
shown in Figure 9,47(b), but with a positive pedestal and reference.
A positive input signal establishes the pedestal level and a triggering
pulse is generated when the ramp crosses the reference level. A
decrease in signal produces',a lower pedestal level and therefore, an
earlier triggering pulse, hence an increase in load voltage. The "gain"
of this type of control can' be expressed in terms of change in load voltage per unit change in signal voltage. For convenience in measurement,
using a rectifier type voltmeter, the load voltage is usually expressed
as the full-wave-rectified average value. Alternate expressions of "gain"
use either the absolute or relative change in signal required to shift the
triggering angle from 150 0 to 30 0 , which represents changing power
in a resistive load from 3% to 97% of full power. The absolute change
in signal level required for this triggering range is the same as the ramp
amplitude. The relative change in signalis the ratio of ramp amplitude
to reference level, usually expressed as a percentage. Since the full
range of power is covered by a smaller range of triggering angles with
inductive loads, the load power factor can change gain upwards by as
much as twice.
Inductive loads, such as induction motors, require a certain logic
in the triggering circuit in order to achieve reasonable symmetry between the positive and negative portions of the alternating voltage.
The PA436 provides this.inductive-Ioad logic by taking the time reference for the ramp-and-pedestal waveform from the zero crossing of
line voltage and by a lock-out gate that prevents trigger pulses from
occurring before the zero crossing of line current.
AC
LINE
FIGURE 9.48
BLOCK DIAGRAM, PA436 PHASE CONTROL Ie
The block diagram of Figure 9,48 shows the functions performed
within the PA436. The DC input signal establishes a pedestal level to
which is added a negative cosine ramp that is derived from the supply
voltage and is externally adjustable. The resulting waveform is com-
268
AC PHASE CONtROL
pared with a zener regulated reference wave in the diHerential comparator which produces an output signal when the ramp is below the
reference level. The lock-out gate blocks this signal from the trigger
pulse generator until after line current has passed through zero and
voltage has appeared across the triac.
IRIDGE
1
MECT'FI£R
TR'r R
,.. _ _ _ _ _ _ _ _ _ 2
REFERENCE
Pt\\\TtL
DIFFERENTIAL
VOLTAG..-
(COMPARATOR
DlYrR~
I
TilliNG
CAPAC'TOR
l
t
REr
- - - - - - - - - - - - 13 - - - - - - - - ,
I
I
I
RBI
I
'"
0"
R,
7K
D. . . .
9
I
I
I
~
~
II
~
~
l~~!...'!!!~~T.!!>~.!!. _______
t
LOCK-OUT
GATE
I
7
c
II
R•
________
...J
t
ztHEft
RAMP GENERATOR REGULATOR
GAIN ADJUST
FIGURE 9.41 CIRCUIT DIAGRAM, PA43I ..PHASE CONTROL IC
The internal circuit of the P A436 is shown in Figure 9.49, along
with a typical external circuit. Operating supply voltage for the circuit
is obtained from the AC line through current-limiting resistor Rs and
the bridge rectifier, and is clamped by the zener diode D9 through transistor Q13 and diode Ds. The clamped waveform appearing between
terminals 1 and 10 is the supply for the pedestal and reference levels.
Note that virtually all circuit current returns through resistor R7 and
diode D 7, and that this current waveshape is a full-wave rectified
sinusoid.
A DC signal, such as from external divider RA and R B , charges
external timing capacitor Cg to the pedestal level through the p-n-p
emitter-follower Q12, supplemented by Qll, with current limited by
R 6 • Capacitor Cg continues charging by a haH-sine-wave current
through QI0 and external emitter resistor Rg , forming the cosine ramp.
This current waveshape is obtained by the voltage drop of supply current through R7, applied to the base of QI0. Amplitude of the ramp
charging current is determined by the external emitter feedback resistor Rg , hence this resistor value establishes ramp amplitude. Diode D7
compensates for the base-emitter voltage of QlO.
The reference voltage level is obtained directly from the zenerclamped supply voltage by divider resistors Rh R2 and Ra. Reference
voltage is brought out on terminal 2 and can be modified, if necessary,
by external resistors to terminals 1 or 10.
269
SCR MANUAL
The differential amplifier Q3, Q4 and Q5, compares capacitor voltage to the reference voltage. The Darlington connection of Q4 and
Q5, in addition to presenting a high impedance to the timing capacitor,
provides an extra base-emitter voltage offset to compensate for the
base-emitter drop of the pedestal emitter-follower Q12. The apparent
reference level (i.e. the voltage required at terminal 12 to trigger at the
beginning of the ramp) only differs from the voltage at terminal 2. by
the relatively small differences in base-emitter voltages of Q3, Q4, Q5'
and Q12.
Common mode current of the differential comparator, through Do
and R4, is controlled by the lock-out gate D 6 , Q7, Qs and Qoo When
load current is flowing through the triac, there is insufficient base drive
on either Qs or Q9 to enable conduction of common-mode current,
hence the comparator is inhibited from producing an output signal to
the trigger. When voltage appears across the triac, current through
external resistor Rr enables the lock-out gate and permits normal functioning of the comparator. The value of Rr determines the triac voltage
required to enable the comparator.
Trigger pulses are generated by the bilateral switch formed by Ql
and Q2 which discharge the external capacitor C 1 into the gate of the
triac. Ql and Q2 are triggered by conduction of Q3, in the comparator,
when the ramp voltage drops below the reference level, but only if
common mode current can flow through the lock-out gate. Since the
trigger pulses alternate with the same polarity as the AC line voltage,
they are ideally suited for triggering triacs directly, or pairs of SCR's
through a 1: 1 pulse transformer.
In order to avoid a carry-over of information from one half-cycle
to the next, the timing capacitor must be reset to a fixed level at the end
of each half-cycle. This reset function is accomplished by Q6 which is
biased off by dividers Rl> R2 and R3 until supply voltage approaches
zero. The capacitor voltage then provides a base drive to Q6, thereby
discharging the capacitor to the base-emitter voltage drop.
9.7.2 Circuit Design With the PA436
Selection of external circuit components is based upon the ratings
and characteristics of the PA436, as follows:
Rs: Minimum value is peak line voltage divided by supply
current peak rating (1 5 -6). Maximum value must supply
sufficient current to obtain zener clamping over desired
triggering range, including current to external loading between terminals 1-10 and 14-10.
C 1 : Must store sufficient charge to trigger the external thyristor.
0.1 p.f will trigger all GE triacs. Peak discharge current
must be limited to pulse rating 13.
R TG : The current limiting resistor (R TG) of 82 ohms is used to
limit the peak trigger pulse output current to its maximum
rating of 15OmA.
R r : Minimum value is peak line voltage divided by enable current peak rating, 10 • Maximum value must supply the
maximum characteristic enable current over the desired
triggering range.
270
AC PHASE CONTROL
Select to produce desired gain from peak sinusoidal ramp
current specification, 113 ramp. Calculate the cosine ramp
amplitude by:
Vrnml'
=(
2 II:! )
wCg
(10,000)
~ volts
To this cosine ramp amplitude there must be added a linear
ramp amplitude which is caused by the comparator darlington base current, 113 bias, where
_ 7 It:!
-3
I
Vrnml' ---C X 10 vo ts
Nonnal range of values for Cg is from 0.1 ILF to 0.01 p.F,
and Rw from 7.5k to lOOk ohms.
Nonn;l range of (RA + RB ) is 10k to 200k ohms. Lower
values can produce excessive loading on the supply. Higher
values limit charging current for C g and cause a peak at the
leading edge of the pedestal that reduces control gain at
the earlier triggering angles. Current gain of the pedestal
emitter follower detennines this effect.
DC Control Signal Source: When a self-contained DC source
is used, such as a tachometer, it should be well filtered and
have an output impedance between 2k and lOOk ohms.
Where a DC supply voltage is needed to create the control
signal, a filter capacitor may be connected between terminals 10 and 14. Loading on this capacitor should be 10k
ohms or higher to minimize charging current. When such a
filter capacitor is used, care should be taken to ensure that
triggering cannot occur before the capacitor is charged to
zener voltage each half-cycle. This can generally be handled
by proper selection of enable current through RI and/or by
adding a small capacitance between terminals 9 and 6 for
a slight phase shift of enable current.
RF Interference Filters: See Chapter 16.
dvldt Suppression Circuits: See Chapter 5.
9.7.3 PA436 in High Power Circuits
When using the P A436 in higher power circuits it is usually necessary to provide a means of gate coupling and gate pulse amplification.
The circuits of Figure 9.50 show five different methods.
For circuits using an SCR-diode pair, Circuit A is the simplest.
The circuit uses SCR2 as a pilot SCR to deliver adequate gate current
to SCRb the main load current SCR. The capacitor C 1 provides a hard
,fire gate signal to allow the circuit to be used for circuits with high
load current di/dt's.
The circuits B, C and D are for inverse parallel SCR's. Since the
PA436 was designed specifically to trigger triacs a triac can in some
cases be used as the pilot SCR. Circuit B shows this type of connection.
The two A14F diodes prevent reverse gate voltage from appearing on
the reverse biased SCR. Circuit C operates in much the same way as B
271
SCR MANUAL
only in C the triac has been replaced by a transfonner with the triac
in its secondary. In this manner the triac circuit can be used regardless
of how high the line voltage becomes.
Circuits D and E use SCR's as the pilot and coupling devices.
Circuit D uses a single SCR as a remote base transistor on the negative
half cycle to provide gate current to SCR2 • Circuit E provides for hard
firing of the load carrying SCR's at the expense of two pilot SCR's and
a pulse transformer.
8211
8211
r------,
r------l
I 6
I
3
PA436
I
3 6 I
I
I
___ I
IL _ _
PA436
I
I
~
Ibl'·------.J
AI4F'S
8211
·-;----3 I
I
I
PA436
8211
I
I
IL _______ .JI
Idl
FIGURE 9.50
TI •
f 3"-6 ....
IL---PA436 I
.... •
1 1
GATING CIRCUITS FOR HIGHER CURRENT SCR's nlGGERED BY THE PA436
9.8 TYPICAL PHASE·CONTROLLED CIRCUITS FOR DC LOADS
Figure 9.51 illustrates the use of SCR's in a typical single phase
center-tap phase-controlled rectifier. By varying R 7 , the DC voltage
across the load can be steplessly adjusted from its maximum value
down to zero. As in the AC phase-controlled switch, a single UJT (Ql)
is used to develop a gate signal to fire both SCR's on alternate halfcycles. Whichever of the two SCR's has positive anode voltage at the
time the gate pulse occurs will fire, thus applying voltage to the load
for the remainder of that half-cycle. The firing angle can be adjusted
by means of R 7 • At 60 Hz, the firing angle of this circuit can be varied
from approximately 10 0 to 180 0 (fully off).
If the secondary voltage applied to the SCR anodes is less than
approximately 100 volts RMS, a separate voltage supply should be
used for the UJT control. In Figure 9.51 an additional 117 VAC wind-
272
AC PHASE CONTROL
ing on T 1 in conjunction with a diode bridge CRc CR 7 can be substituted for CR b CR 2 , and Rl if the main secondary voltage is low. A
more steeply rising square wave of voltage with sufficient amplitude
is thereby provided for control purposes. If the load requires filtering,
inductance Ll and free-wheeling diode CR g may be added, as shown.
L..'~~:.. ..J
CR2
CR,
R,
+
R,-3.3K,5 WATT IF SEC. VOLTAGE OF T, IS
tl7 VOLTS EACH SIDE Of CENTERTAP
R2-47
n,
R3,R4-22
n. '12 WATT
R6 R7
Ra-
lJ2WATT
C,
RS-390 n.112 *TT
_ . _ . _ - 2.7K, I~ WATT
- - - - 0 . 2 MFD
L,-AS REQUIRED FOR FILTERING
FIGURE 9.51
SCR,. SCR2 - AS REQUIRED BY LOAD
SOK LINEAR POT
- - - - - - 3 . 3 K , 5 WATT
CRhCR2.CR4.CR7-G E IN5060
CR,---INI776REGULATING OIOOE
01
G E 2N2646
OPTIONAL CRe-AS REQUIRED BY LOAD CURRENT
PHASE·CONTROLLED DC POWER SUPPLY
When feedback is required a somewhat more elaborate circuit is
required. In order to keep the component count within reasonable limits, the circuit of Figure 9.52 was designed. It uses the PA436 inte-
+
VDC
FIGURE 9.52
IC CONTROLLED REGULATED DC POWER SUPPLY
273
SCR MANUAL
grated circuit phase control which is explained in the previous section,
in conjunction with two SCR's. The feedback is obtained from the
resistor divider of R A , RB , andR 2 • When the voltage at pin twelve (12)
of the PA436 drops below the set point the phase firing .angle is increased to deliver more power to the load. The components can easily
be tailored to almost any voltage and current output desired.
9.8.1 A1.2 KW, 60 VRegulated DC Power Supply
Figure 9.53 shows a regulated DC power supply which utilizes a
cosine modified ramp and pedestal control. Unlike the previous circuits,
this circuit uses two diodes (CRl> CR 2 ) and two SCR's (SCR I , SCR2 )
'to form the main current path. This allows the use of a con-centertapped transformer. By phase controlling the SCR's, control can be
maintained from 7 to 21 volts or 21 to 60 volts by changing the turns
ratio of T s; with a maximum load of 20 amperes in each range.
The function of the feedback element (R1 ), reference (CRg), and
error detector (Qa), all located in the dotted block, is to provide regulation by properly phase controlling the SCR's. The basic operation of
the dotted block is very similar to Figure 9.33. By comparing a portion
of the DC output voltage, as .sampled by the wiper of Rl> with the
stable reference of CR 9 , an error is generated by Qa. Consequently by
adjusting Rl clockwise, the regulated output will increase until it
reaches its maximum value. At maximum output, of course, the system
has no regulating ability. The minimum amplitude is fixed by CRg.
A lower voltage CR g could be used which would allow a lower minimum output voltage.
The dotted block incorporates some interesting features. For example, the cosine-modified ramp and pedestal allow the use of the gain
pot R12 • This should be adjusted for maximum regulation, overshoot,
etc. Also note that the combination of CR 10, Rll and C 4 constitutes a
soft-start circuit. This feature protects the supply when starting under
heavy loads. In addition, stability of the dotted block is achieved by
CR g ,CR9 , the differential amplifier of Ql and Q2,common voltage
points and the unijunction Qa.
Figure 9.54 shows the performance of the supply. The response
times could be reduced if desired at the expense of increased ripple.
274
AC PHASE CONTROL
LI
L2
["----1
.,. I
I
eft3 +
T.
I
T.
I
r
.,
1
C2
CI
cw
I
I
I
I20V
60Hz
I
I
r--------- J
I
T.'
I
••
I
I
L----------- 1
I
••
.,
••
I
I
••
.5
I
.11
.13 I
I
a,
__________________ J
RIG -IOKO{USE FOR 21-60 VOLT RANGE ONLY)
CRI, CA2 - GE MOA
Cia-GE AZBA
R14. RI5 - 330
CR4,CR5,CR6,CR7- GE IN5059
R9- 4.7K.o.
CRB - GE Z4Xll4
CR9- GE FI6HI
RIO- 221<0
RII-22Kn
R12- 1.0 MEGA,1/2W POT
CRIO. CRII- GE AI4F
CI-1300,.fd, 200VDC GE 43F3074CA6
e2- 23,400,Jd.7SVVDC GE 86FI80S
C3 - IOO,.fd,30 WVDC GE WET SLUG 62F403
R3-6800
R4-470KA
RS,R6 ,R7 - 3.3Ktl
C4- 1OO,Jd,25 WVDC
C5- O.I,.f/'v--JV).f'v---<>, { ~~~~T~~:~~~J~WRE
Cr
N
R5ADJ. FOR ZERO NEUTRAL V.
O.Ip.f
n:j~C
_ _ _ _ _ _ _ _ _ _~I~~~-,L___________~
--
j
NOTES;
PHASE B
GAIN 6 REFERENCE AD,J. NOT SHOWN. L -
CRI
-
-
r
__ J
a SCRI RATED FOR LOAD CURRENT
AND SUPPLY VOLTAGE
FIGURE 9.60
PHASE CONTROL OF THREE PHASE CIRCUITS CONTROLLING LINE CURRENTS
D. C. SUPPlY
I
CRI
CR,
I
I
I
!
I
240V L.L.
!~~~~~y <>---t--------------+
60H z
(~~~
INDUCTIVE
LOADI
L
II
'<>A DC
D
I
I
SCRI
I
I
NOTES:
I. CRI, SCRI. CR4 RATED FOR LOAD CURRENT,
SUPPLY VOLTAGE.
2. GAIN
FIGURE 9.61
a
REFERENCE ADJUSTMENTS NOT SHOWN.
THREE PHASE DC POWER SUPPLY USING PA436's
283
SCR MANUAl
power supplies. If for closed loop applications a shorter time constant.
is desired the reference AC frequency should be raised. This will reduce
the needed R-C filter time constant on each supply and yet maintain
the same filtering action.
In order to insure proper· tracking of the control trigger angles
between phases, gain and reference adjustments are needed on two
phases. They are shown on phases A and B as potentiometers Rlll and
R I3 in phase A and R21 and R24 in phase B.
FIGURE 9.62
THREE PHASE FLOATING POWER SUPPLY CONTROL CIRCUIT USJNG P12U .
FOR CONTROL OF PA436'.
The three floating power supplies can be powered from a PA436
and a SC35B triac as shown in Figure 9.63. The circuit usesthePA436
in a single phase mode which provides smooth continuous control of the .
AC voltage supplied to the three floating DC supplies.
c
9
.1,.1
c:r
13
3
6
lOOK
IT
C
TI (WITH AIR GAP)
NOTE:
TI- SECONDARY CIRCUIT
AS IN FIG. 9.62
FIGURE 9.63
284
THREE PHASE FLOATING POWER SUPPLY CONTROL CIRCUIT USING PA436
FOR PHASE CONTROL
AC PHASE CONTROL
There may be applications where it is desirable to common the
control inputs of the PA436's and float only the AC timing line voltages
as shown in Figure 9.64. This would provide rapid response to control
input variations at the expense of necessitating three pulse transformers
and three control transformers to provide the AC timing line voltages.
Figure 9.64(a) as shown is suitable only for resistive loads since
its lockout circuit is not sensing SCR pair voltages. Figures 9.64(b)
and (c) provide for inductive load operation by means of light coupling
between the SCR pair and the PA436's. This is accomplished by means
RI
lOOK
+A
R2
10K
120V
+
t:
•
N
TO PILOT
seRfS
R6
10K
R7
10K
lal
LOAD CI RCUIT
GE
\
2N5778
R2
22K
+
RI
10K
120V
9
R4
22K
SCRI
SCR2
12
INPUT
PA436
N
10
[
[TO PILOT SCR'S
TI
Ibl
GE
2N5778
"" '~II[:
LOAD CIRCUIT
\
R4
22K
leI
FIGURE 9.64
VARIATIONS OF PA436 POLYPHASE CIRCUIT FOR COMMON INPUT
285
SCR MANUAL
of a neon bulb placed across each SCR pair and coupled to the LI4B's
which control the PA436 inductive lockout circuit.
The circuit shown for Figure 9.64(c) is suitable for controlling
transformer primaries. The trigger angle timing is forced to be symmetrical between alternate positive and negative load half cycles
because the PA436 only sees positive half cycles provided by the rectifier bridge. Therefore identical circuitry is used to provide the timing
for alternate trigger pulses. Because of this inherent trigger timing
symmetry no DC voltage should be impressed across the driven transformer primary.
REFERENCES
1. "An All Solid-State Phase Controlled Rectifier System," F. W. Gutzwiller, AlEE Paper 59-217, American Institute of Electrical Engineers, New York, N. Y., 1959.
2. "Phase-Controlling Kilowatts With Silicon Semiconductors," F. W.
Gutzwiller, Control Engineering, May, 1959.
3. "Application of Silicon Controlled Rectifiers in a Transistorized
High-Response DC Servo System," C. Cantor, AlEE CP 60-864,
American Institute of Electrical Engineers, Summer General Meeting, June, 1960.
4. "Speed Controls for Universal Motors," A. A. Adem, General Electric Company, Auburn, N. Y., Application Note 200.47.*
5. "Phase Control of SCR's With Transformer and Other Inductive
AC Loads," F. W. Gutzwiller and J. D. Meng, General Electric
Company, Auburn, N. Y., Application Note 200.31.*
6. "Using the Triac for Control of AC Power," J. H. Galloway, General Electric Company, Auburn, N. Y., Application Note 200.35.*
7. Semiconductor Controlled Rectifiers-Principles and Applications
of p-n-p-n Devices, F. E. Gentry, et al., Prentice-Hall, Inc., Englewood Cliffs, N. J., 1964.
8. "Better Utilization of SCR Capability with AC Inductive Load,"
J. C. Hey, EDN, May, 1966 (also available as reprint from General
Elecb·ic, publication 660.12). *
9. "Solid-State Incandescent Lighting Control," R. W. Fox, General
Electric Company, Auburn, N. Y., Application Note 200.53.*
10. "Transistor. Manual," 7th Edition, General Electric Co., Syracuse,
N. Y.*
*See Chapter 23 for availability and ordering information.
286
MOTOR CONTROLS EMPLOYING PHASE CONTROL
-10
MOTOR CONTROLS EMPLOYING PHASE CONTROL
10.1 INTRODUCTION
Since the AC power line is so universally convenient, and since
phase control is the most convenient way of regulating this power
source, it is little wonder that phase control has been used to control
such a wide variety of motor types. Most of the motors so controlled
however were not designed for this type of operation and were used
because they were available or were low priced. Often the simplicity
of the control circuits is due to a dependence on motor characteristics,
and an improper motor selection will cause poor circuit operation.
Also, even the best control circuit is only part of an overall system, and
can be no more successful than the overall system design.
Most motors are given their ratings based on operation at a single
speed, and depend on this speed for proper cooling. Attempts to use
a motor at a lower speed can cause heating problems. The lubrication
of bearings can also be inadequate for low-speed operation. The presence of odd order harmonics in the phase-controlled wave form, can
produce some odd side effects in induction motors. The speed vs. torque
characteristic of a particular induction motor may make it totally unsuitable for use with a variable voltage control system. Some controls for
universal series motors depend heavily on the existence of a significant
residual magnetism in their magnetic structures, a characteristic that
the motor vendor could be inadvertently trying to minimize.
These potential problems are brought up to a point out the importance of checking with the motor manufacturer to insure that the motor
used is the proper one for this type of use.
The use of a properly chosen and designed motor with a control
of this type can however allow a wide versatility in application. For
instance a temperature compensated furnace blower control can
replace a wide variety of motor sizes and speeds. Now a single, standard
motor can be used with the variable requirements in different installations, being compensated by means of electrical adjustments at the
control. In some cases, where the maximum speed of the motor is set
by the control, the need for designing overvoltage capability into the
motor is eliminated, thus allowing some saving in the motor design.
10.2 BRUSH·TYPE MOTORS CONTROLLED BY BACK EMF FEEDBACK
In order for a circuit to govern the speed of a motor, it must be
able to somehow sense the speed of that motor. The most easily available way to get this information from brush-type motors is by looking
at the back EMF generated by the motor during the time that the
287
SCR MANUAL
controlling SCR is off. In the case of separately excited shunt field
wound, and permanent magnet field motors, this EMF is directly
proportional to speed. In series motors,. the field is not energized at
this time, and residual magnetism must provide the back EMF used
by the circuit. Unfortunately, the residual magnetism is a function of
the past history of motor current, so the voltage the circuit sees is not a
function of speed alone.
Care must also be taken in these circuits that brush noise does not
interfere with circuit operation.
10.2.1 Half·Wave Universal Series Motor Controls
The universal series motor finds use in a wide variety of consumer
and light industrial applications. It is used in blenders, hand tools,
vacuum cleaners, mixers, and in many other places. The control circuits
to be described here can provide the effect of an infinitely variable
tap on the motor.
LOW
UP 10 lAMP
NAMEPLATE
UP TO 3 AMP
NAMEPLATE
MEDIUM
HIGH
UPTOI5AMP
NAMEPLATE
R2
10K IW
IK2W
IK2W
RI
47KI/2W
3.3K2W
3.3K2W
R3
IKII2W
ISOK I/2W
OPTIONAL
I50K 112W
OPTIONAL
10,.f SOY
10,&f 50V
o.lp.f"IOV
OPTIONAL
o.l,.flOV
OPTIONAL
GE
C22BX70
GE
C33B
RI
120VAC
R2>+'-~--~*C~R-I--~~VV--i
GEAI4B
CI
o.5,.f5OV
C2
O.ll£f IOV
SCRI
CR2
GEAI4B
GE
CI06B
Note for 220V. 50/60 Hz Operation: Double value of RI and u.e
at Leost 400 V Semlcon ductor.
(SeR a Diodes' 181
(AI
FIGURE 10.1
UNIVERSIAL SERIES MOTOR CONTROL WITH FEEDBACK
VI
(al
(01
WITHOUT CI
WlTH CI
(el
WITH LARGER CI
FIGURE 10.2 WAYESHAPES FOR FIGURE 10.1
288
MOTOR CONTROLS EMPLOYING PHASE CONTROL
The half-wave circuits of Figures 10.1, 10.3 and lOA supply half
wave DC to the motor. In order to have full-speed operation with these
circuits, the motor must be designed for a nominal voltage of around
80 volts for operation on 120 volt AC lines or 170 volts for operation
at 240 VAC. Brush life of a motor driven by half-wave supply may be
somewhat shorter than for a corresponding motor on full-wave AC ..
The three half-wave circuits shown employ residual back-EMF
feedback to provide increased motor power as the speed of the motor
is reduced by mechanical loading. This back EMF voltage is dependent
on the residual magnetism of the motor which is determined by the
magnetic structure of the motor and the characteristics of the iron.
Care must be taken to ensure that the motor used has sufficient residual
magnetism. For more information see Reference l.
The circuit of Figure 10.1 operates by comparing the residual
back EMF of the motor V2 with a circuit generated reference voltage
V1. If the capacitor C 1 is not present, the voltage V1 is the result of the
divider network composed of R1 and the potentiometer R2. Current
Haws in this branch only during the positive half-cycle due to diode
CR 2. The voltage at V1 then is a half sine wave with a maximum value
at time "A" (Figure 1O.2{a)). If the residual back EMF is greater than
this maximum (the motor is going faster than the selected speed), CR I
will be reverse biased and the SCR will not be triggered and will not
supply power to the motor during this half cycle. As the motor slows
down and its back EMF drops, V2 will become slightly less than VI at
time "A," causing current to How through CR I and the gate of SCRb
thus triggering the SCR. The speed at which CR I conducts occurs may
be varied by adjusting potentiometer R2 which changes the magnitude
of VI. Notice that the smallest impulse of power that can be applied
to the motor is one-quarter cycle, since the latest point in the cycle
that the SCR can trigger is at the peak of the AC line voltage.
If the motor is loaded down so that its speed and back-EMF
continue to drop, the time at which VI becomes greater than V2 comes
earlier in the cycle causing the SCR to trigger earlier, supplying more
power to the motor. If, however, the motor is lightly loaded and running at a low speed, one-quarter cycle power may be enough to change
the motor speed by a considerable amount. If this happens, it may
take a considerable number of cycles to return to the speed at which
the SCR will again trigger. This causes a hunting or "cogging" effect
which is usually accompanied by an objectionable amount of mechanical noise.
In order to alleviate this problem, the smallest increment of power
available must be reduced from a full-quarter cycle to that amount
required to just compensate for the motor energy lost per cycle. To
accomplish this, capacitor C I is added to the circuit. The capacitor
voltage becomes a sinusoid in shape during the positive half cycle.
This voltage is phase shifted by an amount determined by the circuit
time constant ·and an exponential decay during the negative half cycle.
Figure 1O.2(b) shows the results on VI. Two main effects may
be observed. The first is that the latest possible triggering point "A"
is delayed, thereby considerably reducing the smallest increment of
power. The second is that the amount of change of V~ required to go
289
SCR MANUAL
from minimum power to full power, aV, is reduced, providing a more
effective control. Increasing C 1 even more produces. the results of
Figure 10.2(c). It can be seen that the triggering point "A" comes still
later, and a v becomes still smaller. Care must be taken however not
to go too far in this direction, for increasing C 1 decreases av and
increase the loop gain of the system which could lead again to instability and hunting.
It is important that the impedance level of the network formed
by Rh R2 and C 1 be low enough to supply the current required to
trigger the SCRwithout undue loading. It can be seen in Figure 10.2(c)
that this current available for triggering from this network approaches
a sine wave, with its peak at 90°. If the current required to trigger
the SCR is IGT as shown, the latest possible firing point would be at
"B," not at "A" as one would believe from the voltage wave shape.
In many cases, good low-speed operation without a restrictive
specification on gate current to fire would require such a low impedance
network that the power ratings of the resistors and the capacitor size
would become unwieldly and expensive. In such cases, a low-voltage
trigger device such as an SUS can act as a gate amplifier as in Figure
10.3. Use of the SUS in this circuit allows a much higher impedance
network to be used for Rh R2 and C h hence allowing smaller size and
lower cost components. In this circuit the reference voltage V 1 must
exceed the back EMF V2 by the breakover voltage of SUS h which is
about 8 to 10 volts. When SUS I triggers it discharges C 2 into the gate,
supplying a strong pulse of current to trigger SCR I • This eliminates
any need to select SCR's for gate trigger current, and eliminates any
circuit dependence on the trigger current of the particular SCR used.
RI
39K
1/2W
SCRI
120 VAC
CRI
1~~S4~____rG_E_A.14rB__~~~r-______- - J
GE C22B
GE C32B
OR
GE CI22S
112W
CI
0.5,.F
100V
VI
CR2
GE AI4B
FIGURE 10.3
UNIVERSAL
MOTOR
SUS TRIGGERED UNIVERSAL SERIES MOTOR SPEED CONTROL WITH FEEDBACK
Another method of eliminating gate trigger characteristics from
the control's performance is to use a system such as shown in Figure
290
MOTOR CONTROLS EMPLOYING PHASE CONTROL
lOA. Although this circuit also uses the motor counter EMF as a feedback signal, the balance of the system is different. The area enclosed
by the dashed box contains a cosine modified ramp and pedestal circuit
very similar to those described in Section 9.5.2. In this system R4 and
R5 form the pedestal with R2 and Rg providing the ramp current. As
explained in Chapter 4 the Programmable Unijunction Transistor, Ql>
has a variable standoff ratio which is determined by the gate voltage
divider, which, for this circuit, consists of resistors R/l and R 7 •
RI
12K,2W
(25K,4W)
r---
R2
5meg
(10 meg)
I
I
R3
2.5 meg
(5 meg)
R6
2.2K
I
I
120V
(220V)
50/60 Hz
r----------....I
I
I
I
I
I
03
IN5059
(lN5060)
Re
47n
I
I
I
--------,
II
0,
22V
ZENER
R4
5.6K
02
oze06
I
I
I
I
I
I-------------------------~
I
T , . SPRAGUE IIZI2 OR EQUIVALENT
VALUES IN PARENTHESIS FQR 220V OPERATION.
FIGURE 10.4 PROGRAMMABLE UNIJUNCTION TRANSISTOR TRIGGERED UNIVERSAL MOTOR
CONTROL WITH FEEDBACK
The system operates as follows. At the beginning of the positive
half cycle as the line voltage rises the zener blocks current until the
voltage across it reaches 22 volts, at this time the zener clamps the voltage across it to 22 volts. During the first part of the cycle the capacitor
C 1 is charged to a voltage determined by the R4, R5 divider. At the
same time in the gate circuit the voltage on C 2 is building up. When
the voltage on C 2 equals the back EMF plus the forward drop of P3
the diode conducts and clamps the voltage on C 2 to that value. It can
be seen as the motor speed varies this level will change with speed.
When the capacitor C 1 voltage exceeds the gate of Ql then Ql turns
on and triggers the SCR by transferring the charge on C 1 to the SCR
gate through the pulse transformer T 1 • It can be seen that if the speed
is lower than desired the firing angle will advance due to the higher
pedestal and conversely if the speed is too high the triggering angle
will be retarded.
10.2.2 Full·Wave Universal Series Motor Control
Figure 10.5 shows the circuit of a full wave series motor speed
control with feedback which requires that separate connections be
291
SCR MANUAL
available for the motor 'apnature and. field. The full wave bridge supplies power to the series networks of motor field, SCR 1 and armature
Rl and R2. Basically this circuit works on the same principle as that
of Figure 1O.1(a) using the counter EMF of the armature as a feedback signal. When the motor starts' running, the SCR triggers as soon
as the reference voltage across the arm of R2 exceeds the forward drop
of CR 1 and the gate to cathode drop of SCRI. The motor then builds
up speed, and as the back EMF increases, the speed of the motor
adjusts to the setting of R2 in the same manner as the circuit of Figure
10.1(a).
Motor Current
6A
25A
'CRZ
CR3
SCR
C228 or CI228
C328
RI
2K
IW
CR2-5
AI58
A448/A458
CR6 GE AI4B
l
120 VAC
~
R2
5K
2W
CR I
GEAI4B
ARMATURE
CR4
CR5
FIGURE 10.5
FULL WAVE DC CONTROL WITH FEEDBACK
One of the drawbacks of this circuit is that at low speed ,settings,
the anode to cathode voltage of the SCR may not be negative for a
sufficient time.for the-SCR to turn off- because of the decreased back
EMF. When this happens, the motor receives full power for the succeeding half cycle and the motor starts hunting. Furthermore, this
circuit is limited by the fact that SCR 1 cannot be fired consistently
later than 90°. A capacitor on the arm of R2 is not a cure because
there will be no phase shift on the reference due to full wave rectified
charging.
10.2.3.SbuntWeund and P·M Field Motor Control
The shunt~wound DC motor is well suited for use with solid state
speed control systems to provide smooth, wide-range control of speed.
The speed of a shunt motor is inherently reasonably constant with
changes in torque, thus permitting speed control to be achieved by
controlling the voltage applied to the armature. The use of a small
compound series winding can make the speed virtually independent
of torque. Likewise, a small amount of feedback of speed information
292
MOTOR CONTROLS EMPLOYING PHASE CONTROL
into the control that supplies armature voltage will reduce variations
of speed with torque.
GEA40B(4)
BRIDGE
R~R
D3
GEMIB
RI
330K
f
FIELD
SCR
GEC30B
120VAC 60Hz
1
SUS
G!02N4987
FIGURE 10.6 SPEED CONTROL FOR V2 HP, 115 YOLT SHUNT-WOUND DC MOTOR
Figure 10.6 shows a simple and low-cost solid state speed control
for shunt-wound DC motors. This circuit uses a bridge rectifier to
provide full wave rectification of the AC supply. The field winding is
permanently connected across the DC output of the bridge rectifier.
Armature voltage is supplied through the SCR and is controlled by
turning the SCR on at various points in each haH cycle, the SCR turning off only at the end of each haH cycle. Rectifier Ds provides a circulating current path for energy stored in the inductance in the armature
at the time the SCR turns off. Without D:i , the current will circulate
through the SCR and the bridge rectifier thus preventing the SCR
from turning off.
At the beginning of each haH cycle the SCR is in the off-state and
capacitor C 1 starts charging by current How through the armature,
rectifier D 2 , and the adjustable resistor R:\. When the voltage across C 1
reaches the breakover voltage of the SUS trigger diode, a pulse is
applied to the SCR gate, turning the SCR on and applying power to
the armature for the remainder of that half cycle. At the end of each
half cycle, C) is discharged by the triggering of the SUS, resistor R),
and current through R t and R2 • The time required for C 1 to reach
breakover voltage of the SUS governs the phase angle at which the
SCR is turned on and this is controlled by the magnitude of resistor
Ra and the voltage across the SCR. Since the voltage across the SCR
is the output of the bridge rectifier minus the counter EMF across
the armature, the charging of C 1 is partially dependent upon this
counter EMF, hence upon the speed of the motor. If the motor runs
at a slower speed, the counter EMF will be lower and the voltage
applied to the charging circuit will be higher. This decreases the time
required to trigger the SCR, hence increases the power supplied to the
armature and thereby compensates for the loading on the motor.
293
SCR MANUAL
Energy stored in armature inductance will result in the current
How through rectifier D;j for a short time at the beginning of each half
cycle. During this time, the counter EMF of the armature cannot appear
hence the voltage across the SCR is equal to the output voltage of the
bridge rectifier. The length of time required for this current·to die out
and for the counter EMF to appear across the armature is determined
by both speed and armature current. At lower speeds and at higher
armature currents the rectifier Da will remain conducting for a longer
period of time at the beginning of each half cycle. This action also
causes faster charging of capacitor C b hence provides compensation
that is sensitive to both armature current and to motor speed.
This circuit provides a very large range of speed control adjustment. The feedback signal derived from speed and armature current
improves the speed regulation over the inherent characteristics of the
motor.
Another circuit which operates on a similar system is the one of
Figure 10.7. The advantage of this circuit is that for motors whose
field current is less than 4 amperes only four stud mounted semiconductors are needed since diodes D3 and D4 carry only field current.
The second and probably more important point is that these SCR's can
under no condition fail to tum off.
In the circuit SCR 1 and SCR2 .conduct current on alternate half
waves but are triggered from the same trigger circuit. The SCR's therefore carry only one half the current of the SCR in Figure 10.6. Diodes
Dl and D 2 conduct both armature and field current where, as mentioned above, Da and D4 conduct only field current. For currents up
to 1.5 ampere the GE A14B rectifier can be used, for up to 4.5 amperes
the GE A1SB. The advantage of these units is that they are lead
mounted and therefore need no heat sinking except for their tie points.
05
IN5059
(AI4BI
FIELD
RI
Z50K
33K
IZOVAC
SCRz
15K
C
Oz
L -_ _ _ _4-____
.Z,.F
~~
A
NOTE:
FOR 01-04' SCRI. SCRz-, SEE TEXT.
FIGURE 10.7
SCR SHUNT OR PM MOTOR SPEED CONTROL
In this circuit the small amount of feedback which is required for
a shunt motor is the sensing of the armature back EMF. The back EMF
is in the charging path for the capacitor C so the charging is delayed
294
MOTOR CONTROLS EMPLOYING PHASE CONTROL
an amount determined proportional to the back EMF.
Inductance of the field winding of a shunt motor is generally
rather large, resulting in a significant length of time required for the
field current to build up to its normal value after the motor is energized. In general, it is desirable to prevent application of power to the
armature until after the field current has reached approximately normal
value. This sort of soft start function is readily added. For information
on soft starting, see Chapter 9 and Reference 2.
This relatively simple approach is capable of moderately good
speed regulation on the order of 10 percent. For higher performance,
a tachometer feedback circuits as discussed in Section 10.4.3 can be
substituted for the trigger circuit.
10.3 BRUSH·TYPE MOTOR CONTROL-NO FEEDBACK
In many cases speed regulation is not required in the control.
Where the load characteristics are relatively fixed, or where the motor
drive is part of a larger overall servo system, a non-regulating control
circuit may be used. In some cases, these non-regulating circuits can
provide a considerable cost saving over regulated types.
10.3.1 Half·Wave Drive for Universal, Shunt or P·M Motors
Figure 10.8 illustrates one of the simplest and least expensive half
wave circuits. It uses one SCR with a minimum amount of components.
The series network of Rb P b and C 1 supplies a phase shift signal to the
neon bulb which triggers the SCR. Thus, by varying the setting of
potentiometer P b the gate signal of the SCR is phase shifted with
respect to the supply voltage to turn the SCR on at varying times in
the positive AC half cycles. V c fires the neon bulb on both positive
and negative half cycles. The negative half cycles can be disregarded
since both the trigger pulses and the anode voltage of the SCR are
negative.
70 VOLT DC MOTOR ARMATURE
50 VOLT
r- --- --FiELiii
I
I
I
I
I
120VAC
I
I
I
IL
I
GE AI4B
SCRI
GE C22B
OR
GE CI22B
I
I
I
_ _ _ _ _ _ JI
GEAI4B
CONNECTION
FOR SHUNT'
FIELD
FIGURE 10.8
CI
O,II'F
100V
VC
HALF·WAVE CONTROL WITHOUT FEEDBACK (NEON TRIGGERED)
295
SCR MANUAL
By replacing the neon with a trigger device, such as a Diac (the
GE ST2) or a Silicon Unilateral Switch (the GE 2N4987), the performance and reliability of the circuit of Figure 10.8 can be improved considerably because semiconductor trigger devices are longer-lived and
have a more stable triggering point than neon bulbs. Also, becaJIse of
their lower trigger voltage, these solid state trigger devices give a
wider control range. The values of the R-C phase shift network would
have to be increased to compensate for the lower breakover voltages
of these devices.
10.3.2 FUll-Wave AC Drive for Universal Series Motors
Since the universal series motor is generally designed to run on
the 50 or 60 Hz AC lines, the simplest approach to a non-regulating
control is the full wave phase control circuit of Figure 10.9. More
details on the operation of this type of circuit can be found in Chapter 9.
UNIVERSAL MOTOR
(FULL WAVE)
R2
100
RI
250K
TRIAC
(SOOK)
120 VAC1220VAC
I
DIAC
GEST-2
NOTE:
fOR 220/240V, 50/60 Hz R I
FIGURE 10.9
=500K
BASIC NON·REGULATED FULL·WAVE AC PHASE CONTROL FOR UNIVERSAL MOTORS
10.3.3 FUll-Wave DC Motor Drives
SCR's are well suited for supplying both armature power and
field excitation to DC machines.
A full wave reversing control or servo as shown in Figure 10.10
can be designed around two SCR's with common cathode (SCR 2 ,
SCRa) and two SCR's with common anodes (SCR}, SCR 4). In this circuit SCR2 and SCRa are controlled by UJT Ql and the other pair, SCR 1
and SCR 4 , are controlled by UJT Qa. Transistor clamp Q2 synchronizes
the triggering of Qa to the anode voltages across SCR1 and SCR 4 •
Potentiometer Rl can be used to regulate the polarity and the
magnitude of output voltage across the load. With Rl at its center
position, neither UJT triggers and no output voltage appears across
the load. As the arm of Rl is moved to the left, Ql and its associated
SCR's begin to trigger. At the extreme left-hand position of Rl> full
output voltage appears across the load. As the arm of Rl is moved to
the right of center, similar action occurs except the polarity across the
load is reversed.
296
MOTOR CONTROLS EMPLOYING' PHASE CONTROL
If the load is a DC motor, plugging action occurs if Rl is reversed
abruptly. R14 and R l5 are used in series with each end of the transformer to limit fault current in the event a voltage transient should
trigger an odd- or even-numbered SCR pair simultaneously. Commutating reactor T 3 and capacitor C s limit the dv/ dt which one pair
of SCR's can impress upon the opposite pair.
",0
"..
+20V.
OUTPUT VOLTAGE
CONTROL
",
AC
SUPPLY
".
"2
",
T,
ov.
"I.
-3V.
",-lOOK LINEAR POT
Rz."a-470 OHMS,I/2 WATT
"3,"9-2700 QtfMS,I/2WATT
R4,Re-IOK,2WATTS
"5-4700 OHMS,IIZ WATT
R I I , R I 2 - - - - 2 2 0 0 OHMS,ZWATTS
CRI,CRZ.CR3,CR4,CR5,CRe--GE IN~60
CRr
Gf AI4F
"14,".,
~I:M~NS:T,=!T~~~~~SS.DEPEN-
Q',Q3
02
Of 2N1671A
Of GET 2222 .
CI,tZ
0.2 MFD
T"T2
P£223I, SPRAGUE IIZI3
SCRI.SCR2.SCR3~CR4-:NT~: ~~~~r::AER:A:~~~~~~~-
~:}
AS REQUIRED BY lOAD
OR EQUIVALENT
MER VOLTAGE)
FIGURE 10.10 FULL WA¥E REVERSING DRIVE
10.3.4 Balanced·Bridge Reversing Servo Drive
A phase-sensitive servo drive supplying· reversible half· wave
power to the armature of a small permanent magnet or shunt motor is
shown in Figure 10.11. The power circuit consists of two half-wave
circuitsback-to-back (SCRb CRl; and SCR2 , CR2 ) triggered, by unijunction transistor, Qh on either the positive or negative half-cycle of
line voltage depending on the direction of unbalance of the reference
bridge resulting from the value of the sensing element R I . RI can be a
photo-resistor, a thermistor, a potentiometer, or an output from a
control amplifier.
The potentiometer Rs is set so that the DC bias on the emitter of
unijunction transistor, Qh is slightly below the peak-point voltage at
which QI triggers, by an amourit dependent upon the deadband
desired. With RI equal to R2 the bridge will be balanced, UJT (QI)
will not trigger and no output voltage appears across the load. If Rl
is increased thus unbalancing the bridge, AC signal will appear at the
emitter of the UJT causing the emitter to be biased above the trigger
voltage during one half-cycle of the AC. Ql will trigger and, since
SCR2 is forward biased, SCR 2 will trigger. When Rl is decreased,
297
SCR MANUAL
similar action occurs except that SCR 1 will trigger, reversing the
polarity across the load.
CR7
120VAC
M~+-~--~--------------------~
MOTOR
ARMATURE
CRI,CR2,SCRI,SCR2: AS REQUIRED BY LOAD
RI : 3.3K bE NOTE R6: IK
CI : O.lpf
R2: UKJ""
R7: 47n
QI: GE2N2646
R3: 3.3K,2W
R8,2500n
CR4:GEINI776
R4: 3.3K,2W
R9: 4711
CR5-8:GEIN5060
R5: 2 MEGOHMS
RIO: 470
CR3: GE DT230A
NOTE'EITHER RI OR R2 MAY BE A VARIABLE RESISTANCE TRANSDUCER,
SUCH AS THE GE 8425B PHOTOCONDUCTOR, OR A .
THERMISTOR, OR A POSITION SENSING POTENnoMETER.
FIGURE 10.11
BALANCED-BRIDGE REVERSING SERVO DRIVE FOR SHUNT-WOUND MOTORS
-If instead of a shunt wound motor, the servo motor is series wound,
a circuit similar to Figure 10.12 can be used. In this circuit the triac
is tr~ggered only on either the positive half wave or the negative half
wave. Since the armature is within a bridge the armature voltage is of
one polarity regardless of which half cycle the triac is energized. The
field winding current, on the other hand, is reversed with a change
of triac triggering polarity. The triac control circuit features control of
.all the control parameters, gain, balance anddeadband and also provides for an analog control voltage input. If desired the balance pot
can be replaced by a pair of resistive transducers or a positioning
potentiometer.
10.4 INDUCTION MOTOR CONTROLS
There are a wide variety of induction motor types and within
these a wide range of possible characteristics. Some of these characteristics can make a. given motor type unsuited for control by means
of phase control. The most. obvious difficulty is that induction motors
tend to be more frequency sensitive than voltage sensitive, while phase
control generates a variable-voltage, constant-frequency source. If the
motor was not designed for use with phase control, the motor designer
may have accentuated this problem' in order to get better speed regu-
298
MOTOR CONTROLS EMPLOYING PHASE CONTROL
SERIES} REVERSIBLE
\\ FIELD
~~~h~S
CR4
R2
5600
120 V.
60Hz
R3
5600
ARMATURE
CR5
R4 -25K
"BALANCE"
CR6
+
CIIO.f
CRI
GEDT230B
TRIAC
CR2
GEDT230B
R5
100
+
OUTPUT
VOLTAGE
_----?......j...L-,.....----++
CONTROL VOLTAGE
Notor Current
Diodes CR3 -CRS
Triac
1.5A
4.5A
AI4B
AI58
SCI41B
SCI41B
4.5A
A40B I A41B
SC51B
FIGURE 10.12
BALANCEO BRIDGE REVERSING SERVO DRIVE FOR SERIES·WOUND MOTOR
lation. In general, the motor used should be as voltage sensitive as
possible. Variable voltage drive of induction motors is a compromise,
one usually dictated by economics but very satisfactory in properly
implemented applications. A variable frequency drive would be superior in some applications, but generally far more expensive than phase
control. Information on variable frequency inverters that can be used
for drives can be found in Chapter II.
Certain types of single phase induction motors, notably split-phase
and capacitor start motors, require a switched start winding. Since
there is a torque discontinuity when the start switch cuts in or drops
out, it would be impossible to control the speed of the motor around
these points. This means that where the higher starting torque of a
switched start winding is required, the motor ,should be designed so
299
SCR MANUAL
that the switching point is below the range of speed ov.er which phase
control is desired.
Another important consideration is the power factor of the·motor.
An overly inductive motor can require a fair degree of sophistication
in the control in order to avoid the problems. associated with phase
control of inductive loads. This topic is. covered in detail in Section 9.6.
10.4;1 Non-Feedback Controls
Unlike brush type motors, induction motors give no convenient
electrical indication of their mechanical speed. This means that direct
speed feedback is not nearly as easily available. For some applications
like fixed fan loads, direct voltage adjustment with no feedback yields
satisfactory performance. An example is the circuit of Figure 10.9
when working with a perman~nt split capacitor motor or with a shaded
pole motor. The need for the proper motor-load combination is shown
by the speed torque curves of Figure lO.13. In the case of a low rotor
resistance, Figure 10.13(a), it can be seen that varying the voltage of
this motor will produce very little speed variation, while the higher
rotor resistance motor of Figure 10.13(b) would give satisfactory results.
S.
SPEED SIT_-_""_~--/
52
TORQUE( Al POOR - LOW ROTOR RESISTANCE
FIGURE 10.13
TORQUE(B) GOOD - HIGH ROTOR RESISTANCE
INDUCTION MOTOR SPEED-TORQUE CURVES FOR USE WITH'A FAN-TYPE LOAD
10.4.2 Indirect Feedback
Often the problem of speed regulation of induction motors may
be bypassed by considering the complete system control problem. As
an example, consider the problem of controlling the speed of the
blower in a hot air' heating system in response to the temperature of
the air. It can be seen that what is of prime interest is the temperature
of the air, not the precise' speed of the motor. This kind of analysis
can lead to the circuit of Figure 10.14.
In this circuit, thermistor Rs acts in response to air temperature
to control the power supplied to the motor. Resistor Rl and its phase
control network serve to set a minimum blower speed, to provide con300
MOTOR CONTROLS EMPLOYING PHASE CONTROL
THERMISTOR
R3
RZ
100
IZOV
60Hz
C3
QI"fd
FIGURE 10.14 FURNACE BLOWER CONTROL
tinuous air circulation and to maintain motor bearing lubrication.
Figure 10.15 gives an example of a more sophisticated control
system, capable of a much higher control gain. This could be used to
control a blower motor in response to room temperature for heating
control, or in response to cooling coil temperature to prevent air conditioner freeze-up.
DI
C4
IlOV
60Hz
I
L
DI-D4: (4) GE AI4B
OR (I) GE BI02
05: GE AI4 A
06: GE INI776
01 : GE 2NZ646
CI BI C2 : 0.11". 50V
R I : 6800n 2W
R2 : 470Kn 112W
R3:5MEGII2W
R4 : IKn 112W
R5 : 10Kn IW
R6 : SEE NOTE
R7
C3
TI
C4
R8
33n 112 W
0.02,.1. 20DV
XFMR 1:1
SPRAGUE IIZI2 OR EOUIV.
0.11' f 400V
470Kn 112 W
NOTE: IN THE ABOVE ARR·ANGEMENT CIRCUIT IS SET UP FOR A HEATING APPLICATION. IN A CODLING
APPLICATION R6 BI. R5 ARE INTERCHANGED. R6 SHOULD BE A THERMISTOR WHICH WILL AFFORD
3Kn TO 5Kn AT TEMPERATURE DESIRED. 'TEMP ADJ" R5 SHOULD BE SET UP TO PROVIDE FULL
''oN'' AT DESIRED UPPER TEMP. OF THE THERMISTOR R6. 'GAIN" R3 OR 'BANDWIDTH' MAY THEN BE
SET FOR "FULL OFF' (ZERO SPEED) CONDITION AT DESIRED 'LOWER TEMP' OF R6.
FIGURE 10.15 TEMPERATURE CONTROL OF SPEED; SHADED POLE AND TSC MOTORS
This is a ramp-and-pedestal system designed for the control of
fan or blower motors of the shaded pole or permanent split capacitor
301
SCRMANUAl
type in response to temperature of a thermistor. The circuit includes
RF noise suppression and dv/dt suppression.
10.4.3 Speed Regulating Control of Induction Motors
In order to actually regulate the speed of an AC induction motor
by means of phase control, it is necessary to provide speed information
to the circuit by means of a small tachometer generator. Such a generator could be made quite inexpensively, as high precision is not necessarily required. In addition, the speed torque characteristics of the
motor should be quite voltage sensitive such as that of Figure 10.13(b).
A motor such as that of Figure 10.13(a) would be quite difficult to
control in astable manner, as the open loop system characteristics
would be highly nonlinear through the controlled speed range. The
drop-out point of the start switch, if present, should be below the
lowest desired controlled speed.
Figure 10.16 shows a general block diagram of such a control
system. For a practical circuit, a ramp and pedestal control circuit,
with inductive load consideration (as shown in Figure 9.35) can be
combined with the input connection shown in Figure 10.17. This con-
FIGURE 10.16 BLOCK DIAGRAM OF AN INDUCTION MOTOR SPEED CONTROL SYSTEM
nection is shown for an AC tachometer in the 4 to 6 volt range. The
R 1-C 2 time constant is chosen to give adequate filtering at the lowest
desired speed and tachometer frequency, consistent with system sta-
,
07
01
03
TACH
02
C2
RI
04
FIGURE 10.17 AC TACHOMETER CONNECTION TO A RAMP AND PEDESTAL TRIGGERING CIRCUIT
302
MOTOR CONTROLS EMPLOYING PHASE CONTROL
.1150
INDUCTION
I
MOTOR
0
AC
TACH
4.7K
10K
2W
L..-_-+
220K
r---~~~r-+--1~9
120VAC
.05/50
100
82n
.11200
SPEED
NOTE:
WITH VALUES SHOWN. SPE£D .RANGE IS ABOUT 600 TO 1800 RPM USING AN 8 POLE TACH.
FIGURE 10.18
INDUCTION MOTOR SPEED CONTROL USING THE PA4.36 AND A FREQUENCY
DEPENDENT TACHOMETER CIRCUIT
bility requirements. This system is also applicable to multiphase controls as wen as DC motor drives.
For higher performance systems the PA436 can be used. Either
ac or dc tachometers can give adequate speed control. Figure 10.18
shows an induction motor speed control using an ac tachometer. Each
time theac tachometer output swings positive, C 1 and C 2 are charged,
and on the negative swing C 1 is discharged. The speed control potentiometer controls the discharge of C 2 and hence the apparent feedback
voltage presented to pin 12 of the PA436 Section 9.7 can be referred
to for details of the P A436
Figure 10.19 shows two other tachometer connections that can
be used with the PA436 . In these connections it is important that
the output voltage of the tachometer is proportional to speed and has
a voltage of at least 4 volts at minimum speed.
RIPPLE
FILTER
12O-_ _41V1r..,
o
12o---_--_-'VII'Ir'"-<:'
IOo---~--~------~
__~~
Al D.C. TACHOMETER
FIGURE 10.19
IOo-~-4--------~-4--~----~
Bl A.C. TACHOMETER USED AS
D.C. TACHOMETER.
OTHER PA436 TACHOMETER INPUTS
10.5 SOME OTHER MOTOR CONTROL POSSIBILITIES
In addition to controlling or varying the speed of a motor, there
are several other control functions which can be done using solid-state
control.
303
SCR MANUAL.
One of the simplest functions is the use of a triac as a static switch
for contactor replacement. When used with a reversing-type permanent
split capacitor motor, a pair of triaes can provide a rapidly responding,
reversing motor control, as shown in Figure 10.20. S1 and S:! can be
c
120YAC
RI
2I150W
R2
R3
10011
112W
112W
10011
51
FORWARO
52
REVERSE
FIGURE 10.20 T1lIAC CONT1l0L OF A REVERSIBLE PERMANENT SPLIT CAPACITOR MOTOR
reed switches, or the triacs can be gated by any of a number of other
methods. Also, use of a solid-state static switching circuit, with an
appropriate triggering circuit, can provide a motor overtemperature
control which senses motor winding temperature directly.
In this circuit it is important to insure that the triacs have sufficient voltage rating. It can be seen that if one triac is on the other triac
must block voltage that is greater than line voltage due to the L-C
ring between the capacitor and the motor winding. As a rule of thumb
the triac voltage rating should be 1.5 times the capacitor's voltage
rating.
The second point regarding Figure 10.20 is that Rl be sized correctly to insure that if one triac is switched on when the other is conducting, the surge current from the capacitor discharge is limited to a
safe value.
10.5.1 Single·Phase Induction Motor Starters
In many cases, a capacitor start or a split-phase motor must operate where there is a high frequency of starts, or where arcing of the
mechanical start switch is· undesirable, such as where explosive fumes
could be present in the neighborhood of the motor. In such cases, the
mechanical start switch may be replaced by a triac. The gating and
dropout information may be given to the triac in several ways.
Perhaps the simplest form ·of connection is to use a conventional
current or voltage sensitive starting relay as pilot contacts for a simple
triac static switch as shown in Figure 8.1(a).
304
· MOTOR CONTROLS EMPLOYING PHASE CONTROL
Another method is that shown in Figure 10.21 which shows the
triac gated on by the motor current through a small current transformer. As the motor speeds up, the current drops off and no longer
trigger the triac. A variation of this is to replace the current transformer
with a small pickup coil, which is mounted near the end windings of
the motor. This gives a somewhat more precise signal for .the triac.
RUN
WINOINe
r-----
DISCHA~
AC
SUPPLY
RESISTOR
(OPTIONALl'- ____ _
TRIAC
CURRENT ~~~~-I
TRANSFORMER
FIGURE 10.21
TRIAC MOTOR STARTING SWITCH
Where a tachometer generator is already sensing the speed of
the motor, this same signal can be used to control a triggering circuit
for the triac. With this arrangement the dropout· speed can be precisely
set, so as to be outside the desired speed control range.
REFERENCES
1. "Speed Control for Universal Motors," A.A. Adem, General Electric
Company, Auburn, New York, Application Note 200.47.*
2. "Speed Control for Shunt-Wound DC Motors," E. Keith Howell,
General Electric Company, Auburn, New York, Application Note
200.44.*
3. "Phase Control of SCR's With Transformer and Other Inductive
AC Loads," ·F. W. Gutzwiller and J. D. Meng, General Electric
Company, Auburn, New York, Application Note 200.31.*
4. "Using the Triac for Control of AC Power," J. H. Galloway, General
Electric Company, Auburn, New York, Application Note 200.35.*
·Refer to Chapter 23 for availability and ordering information.
305
SCR MANUAL
NOTES
306
ZERO VOLTAGE SWITCHING
11
ZERO VOLTAGE SWITCHING
11.1 INTRODUCTION
When a power circuit is switched "on" and "off", high frequency
components are generated that can cause interference problems (see
also Chapter 17). When power is initially applied, a step function of
voltage is applied to the circuit which causes a shock excitation. Random switch opening chops current off, again generating high frequencies. In addition, abrupt current interruption in an inductive circuit
can lead to high induced voltage transients.
The latching characteristics of thyristors are ideal for eliminating
interference problems due to current interruption since these devices
can only turn off when the on-state current approaches zero, regardless
of load power factor.
On the other hand, interference free tum-on with thyristors
requires special trigger circuits. It has been proven experimentally that
general purpose AC circuits will generate minimum electromagnetic
interference (EMI) if energized at zero voltage.
The ideal AC circuit switch therefore consists of a contact which
closes at the instant when voltage across it is zero, and opens at the
instant when current through it is zero. This has become known as
"zero voltage switching."
Zero voltage switching is not new. First proposed in 19591, it is
rapidly gaining considerable acceptance, particularly with regard to
electrical heating applications. 2 ,3 This chapter will consider both discrete and monolithic integrated zero voltage switching circuits, its
attendant benefits and problems, and potential uses of zero voltage
switching at frequencies higher than 50 or 60 Hz.
11.2 ELECTROMAGNETIC INTERFERENCE
Each time a circuit is energized or de-energized, one must be
concerned with the electrical disturbance which this may cause. 4 Each
time a thyristor energizes a resistive circuit; load current goes from
zero to the load limited current value in a few microseconds. The frequency analysis of such wave forms shows an infinite spectrum of
energy in which the amplitude is inversely proportional to frequency.
In applications where phase control is used, the AM broadcast band
would suffer severe interference, for example, with less problems with
TV and FM as shown in Figure 1l.l. This curve shows a plot of quasipeak microvolts of conducted interference. This is one of two basic
. types of radio frequency interference. In addition to that which is conducted through the power lines, there is the question of radiation from
the circuit itself. This can be minimized by keeping the physical size
307
SCR MANUAL
of the current loops formed by the thyristors and the EMI filter network to a minimum. In addition to these two forms of radiation, there
are also' questions copceming telephone interference and acoustical
noise. The uppermost curve in Figure 11.1 is for a typical unsuppressed
600 watt lamp dimmer design using thyristor switches. Notice also that
a curve is given for a typical food mixer and for a noisy 40 watt fluorescent lamp. Thus it can be seen that thyristor control is not alone in
producing interference in the AM band, while for FM and TV, the
thyristor interference is negligible compared with the other conventional components.
IVOLT.-------------.-------------~----------__.
IOO,OOO~------~----~------------+-------------~
tl5V (RMSl
60Hz
'"~
~
:i
~
~
~
~
ffi
II:
...au~
o
au
10,000 ~----------+H~~--------+---------------i
TYPICAL
FOOOM"
~
,
8
QUIET
'.....
40WATT../-,
FLOURESCENT ,
,
10.0r----
,
1",
f,
lAM
IeROAOCAST'
IBAND
I
10~
____
100KHz
•I• I •,I
~_L-_L
_________1 __ _ _ _ _ _ _ _ _ _
I MHz
10MHz
~
IOOMHz
FREQUENCY
FIGURE 11.1 CONDUCTEO INTERFERENCE FROM SEVERAL SOLID STATE POWER
SWITCHING CIRCUITS"
308
ZERO VOLTAGE SWITCHING
The "suppressed" dimmer has an LC filter network to slow the
rate of rise of current and absorb the higher frequencies. Chapter 17
details the design of these filters. It can be seen that the simple LC
filter just manages to meet the NEMA WD-2 limit and that a single
inductor would not. For large loads, these filters not only become bulky
and expensive but they also will dissipate much power, and zero voltage
switching becomes a more and more attractive alternative.
Figure 11.2 compares the behavior of the same LC suppressed
phase control circuit with two synchronously switched circuits. At
1 MHz, the noise figure of the synchronously switched circuits is an
order of magnitude lower than the filtered phase control circuit.
10,000
~
"
,ICAl SUPPRESsio iHjj FiiliRDllEO
'"
TRlACS OR
~=~f
<5VANODE
,
NEMA WO-2 LIMIT
'\.
VOLTAGE
1\
TRIACS OR
ANTI-PARALLEL
SCR's CONTINUOUSLY
GATED WITH DC
11
~
""""I~
,'-.'\.
I
10
0.1
100
FREQUENCY (MHz)
FIGURE 11.2 COMPARISON OF NOISE PERFORMANCE OF FILTERED PHASE CONTROL
CIRCUIT .. ZERO VOLTAGE SWITCHING CIRCUITS"
It is interesting to note the superior performance of the continuomily gated triac circuit at the lower frequencies. The DC gating signal
assures that the triac (or SCR) is always "on", that it does not unlatch
at zero current due to an unsufficient holding current. The DC gating
signal would have less interference than a pulsed gate signal if the
pulse occurs when the anode voltage is greater than 5 volts.
Figure 11.2 also shows that the criterion to meet the NEMA WD-2
interference limit is that the triac or SCR must be turned on before
the line voltage rises above 5 volts. Table 11.1 shows the times and
angles for 5 volts for diHerent voltages and frequencies.
309
SCR MANUAL
Voltage (RMS)
115
24
frequuCJ
(Hz)
50
60
400
TABLE 11.1
220
9(")
T (asec)
9
T
8.47
8.47
8.47
471
392
58.8
1.76
1.76
1.76
97.9
81.6
12.2
T
9
.92
.92
.92
51.2
42.6
6.4
THIS TABLE SHOWS THE ANBLE AND'THE TIME AT WHICH POINT THE
POWER VOLTAGE EXCEEDS 5 VOLTS
.
11.3 DISCRETE ZERO VOLTAGE SWITCHING CIRCUITS
Discrete zero voltage triggering circuits abound so that the intent
of this section is to show some of the more typical ones, how they work,
what they can do, their limitations, etc. The important idea is to
ensure that the thyristor turns on before the instantaneous voltage
across it exceeds 5 volts in order to meet the NEMA WD-2 EM11imits.
11.3.1 Basic Switching Circuit
The circuit shown in Figure 11.3 accomplishes ideal switching for
a half-wave circuit.6 Other variations (including full wave) will be
discussed later.
LOAD
LINE
D,
DT230B
R,
2.2K
VOLTAGE
D3
D1230S
R3
220K
SCR,
C'OIS
'20V
SO/60Hz
RS
'OK
.~
Q,
D2
vvV
gc:"
r<:'R
~AGEB~A~B
CCNTROL
CONTROL
SWITCH
OPENS
SWITCH
CLOSES
RANDOMLY
RANDOMLY
2N5I72
47K
0000
VOLTAGE
ACROSS C,
t vv,
·DT23OF
(a> Ideal Half Waye Switching Circuit
(b) Associated Voltage Half.Wave forms
FIGURE 11.3 HALf WAVE ZERO VOLTAGE SWITCHINB CIRCUIT
When transistor Ql is cut off, positive anode voltage on SCR I
causes gate current to How through Ds and R4, triggering SCR I into
conduction. When transistor Ql is biased into conduction, current
through R4 is shunted away from the gate of SCR I through the collector of Ql.
The contacts of switch SI (or the value of resistance connected
across its contacts) control the conduction state of Ql. With the contacts open, the negative half-cycle of the supply charges capacitor C l
to the peak of the supply voltage through Rl and D 1 • As the AC supply
voltage drops from its negative peak, capacitor C l discharges through
D2 and R2 , thus applying a cutoff bias to Qt. This causes SCR 1 to
310
ZERO VOLTAGE SWITCHING
trigger as soon as the AC supply voltage swings positive through zero,
thus providing synchronous closing. The SCR's latching characteristics
maintain it in the conducting state for the remainder of the positive
half cycle and open the circuit syllchronously at the point where load
current reaches zero naturally. Although the contacts of switch SI are
open during the random interval A indicated in Figure 11.3(b), the
SCR conducts only for complete half cycles.
If the control contacts are closed, capacitor C 1 does not charge
during the negative half cycle. Ql is therefore driven into saturation at
the beginning of the positive half cycle before SCR 1 can be triggered,
and gate current is shunted away from SCR 1 for the remainder of that
cycle regardless of subsequent switching of the control contacts during
the positive half cycle. Also no bypass resistor (RGK ) is required for
sensitive gate SCR's since the gate is shorted by the transistor. The
following design criteria must be followed for the circuit of Figure 11.3(a):
I) The resistance of R4 must be less than the line voltage at which
switching should occur (typically 3 to 5 volts) divided by the
maximum gate current required to trigger the SCR. This latter
requirement highlights the desirability of an SCR with sensitive gate triggering characteristics.
3V
3V
R4
= 1;;=
200pa "'" 15Kn
The lower limit of R4 is determined by the collector current
limit of Q1.
2) R;i in tum must provide sufficient base drive to Ql to keep Ql
in saturation throughout the cycle when C 1 is in the discharged
state. Using 15 as a conservative figure for the current gain of
the 2N5172:
Ra
= 15 . R4 "'" 220 K
3) R2 should be substantially less than Ra. Pick
R2
47Kn
4) The time constant of R2C 1 must be sufficient to extend bias
current for Ql into the positive half cycle, and hence should
be approximately lhf. For 60 Hz:
8.3 msec
R2C 1
C 8.3 msec ""'.2 ..f
1 47 K
r
5) Resistor Rr;, which limits the capacitor discharge current
through the contacts when Sl is closed, must be low compared to R J • This will prevent C 1 from charging when the control contacts are closed.
PickR"
10K
=
=
=
11.3.2 Two Transistor Switching Circuit
Figure 11.4 is an extension of Figure 11.3 in that Ql along with S
provides the gating signal while Q2 detects the zero voltage crossing.
Switch S could be connected between base and emitter of Ql to provide inverse logic, i.e., SCR 1 off with S closed.
311
SCR MANUAL
Al4B
10Vo-------<~---_.
100
115V
SO/60Hz
°2
OT230F
S
°1
SCR I
CI22B
DT230F
FlaURE HA TWO TRANSISTOR ZERO yolTAaE SWITCHINa CIRCUIT. DOTTED PORTION
AT RlaHT FOR I"TEaRAL CYCLE CONTROL
Diodes Dl and D2 perform the same function as D2 and D4 of
Figure 11.3, namely protection of low voltage components during the
negative half cycle. The forward junction voltage drop of D2 ensures
that SCR1 is not triggered on while either Ql or Q2 are on.
The same rule applies to specify Rl as shown in Step 2 of Section
11.2. Again the lower limit of Rl is determined by the allowable base
current of Q2. The use of darlington transistors for Qt and Q2 would
allow larger values for Rl and R4 but this would also necessitate an
additional diode in the gate circuit to compensate for the higher saturation voltage of this type of transistor.
The dotted in connections are for integral cycle control. SCR2 is
slaved to SCR 1 by R2, Rs and C (see Section 8.3). The elimination of
any possible half-waving prevents saturation effects even in critical
elements like transformers having marginally designed magnetic cores.
11.3.3 CSCR Zero Voltage Switch
Another useful zero voltage switch can be easily assembled using
a complementary or n-gate SCR, such as the CI3Y. H switch S of
Figure 11.5 is closed, the C13Y can be turned on by gate current How
through resistors Rh R2 and diode D 1 • However, as soon as the line
voltage rises above 5 volts, diode D1 becomes reversed bias and the
C13Y can no longer turn-on. Since the gate of the C13Y is sampling
the SCR anode voltage, this circuit may be used with any load power
factor. Resistor Rs should be chosen so that the leakage current through
Dl does not damage the gate of the C13Y.
V GR11 + 5V
(11.1)
Rs<
IR
where:
VGRM
C13Y gate avalanche voltage, typically 5 volts
IR = reverse leakage current of D t .
=
312
ZERO VOLTAGE SWITCHING
AI58
LOAD
01
0T230B
5V~
150
RI
\I 5V
50/6 OHz
R2
lOOK
~ Q:y
(; ~C122B
R3
FIGURE 11.5 CSCR ZERO VOLTAGE SWITCHING CIRCUIT
This circuit is easily scaled up for 220 volt operation.
11.3.4 Triac Zero Voltage Switching Circuits
In Figure 1l.6, the triac will be gated on at the start of the positive half cycle by current How through the 3 pf capacitor as long as
the C103 SCR is. off. The load voltage then charges up the 1 pf capacitor so that the triac will again be energized during the subsequent
'negatiye haH cycle of line voltage. Note that a selected gate triac wilJ
be required because of the 111+ triggering mode (see Chapter 7.)
1.2K
lOW
3,.F
AI4B
15011
2W
115V
AI4B
TRIAC
AI4B
50/60Hz
CI03B
I,.F
200V
TRIGGER
IK
IK
IW
FIGURE 11.G TRIAC ZERO VOLTAGE SWITCH
Zero point switching is assured by the SCR. A change state of the
CI03 from "off" to "on" during the positive half-cycle will have no
effect on the triac since it will already be latched on. Furthermore, if
the CI03 is turned on at the start of the cycle, the triac cannot be
triggered at any time during that cycle since the C103 wilJ say on
until reverse biased.
313
SCR MANUAL
The major difficulty encountered with this circuit involves triggering the triac during the positive half cycle. Because of the high gating
current requirements of the triac, line voltage often reaches 10-15 volts
before the triac fires ..Larger capacitors and smaller resistances would
advance this firing angle but would also increase the power dissipation
in the gate, in the C103 and in the components themselves.
11.3.5 Improved Zero Voltage Triac Switches
Since the problem of triggering the triac early in the cycle occurs
only for the positive half-cycle (the 1 pi capacitor applies DC to the
gate of the triac during the next zero voltage crossing so that the triac
does not commutate off), one solution is pilot triggering with a sensitive gate SCR.
In Figure 11.7, the pilot C106 SCR is turned on very shortly after
the voltage starts to rise by Rl assuming the C103 SCR is off, which in
tum triggers the triac. Negative half cycle triggering occurs as before.
The maximum voltage to trigger the C106B is:
VM
IGT(max) • Rl + 4 . VF
where
VF
PN junction drops
"'" (200 p.a) (10 K) + (4) (.6 V)
=
=
= 4.4 V
Diode Dl is required in order to prevent the 1 pi capacitor from
charging negatively during the negative half cycle when the triac is on.
If it does, it triggers the pilot SCR when the C103 has just turned on.
CI06B
RI
10K
DT230B
TRIAC
DT230B
115V
50/60 Hz
CI03B
DT230B
TRIGGER
IK
IK
DT230B
DI
FIGURE 11.7
I"F
IMPROVED ZERO VOLTAGE TRIAC SWITCHING CIRCUIT
Both the circuits of Figures 11.6 and 11.7 require selected gate
triacs. Figure 11.8 allows the use of the standard type since the gate
modes are now 1-, 111-, i.e., negative gate triggering.
314
ZERO VOLTAGE SWITCHING
0T230B
TRIAC
115V
50/60Hz
10K
FIGURE 11.8 ZERO VOLTAGE TRIAC CIRCUIT USING STANDARD TYPE TRIACS
11.3.6 Transistorized Zero Voltage Trigger
Figure 11.9 shows a zero voltage triggering circuit that also includes its own regulated power supply. The circuit operation is as
follows:
1) The power supply capacitor C 1 is charged up to the zener
voltage of D6 through diode Dij and R 2 , typically 6-7 volts.
02
R6
TRIAC
R5
04
Q3
IISV(220V)
60Hz
Q2
L
0
A
0
CONTROL
INPUT
R2
01'04'
05'
06'
QI •
Q2'Q4 •
OH0805
OT230F
E-B OF 2N5172
2N5354
2N5172
R,-R4
R2
R3
R5
R6
8.2K
• 10K,2W*
= 4.7K
·33
=3.3K
:I:
C,
:l:IOO,..F,IOV
FOR LIGHT ACTIVATION, USE
2N5777 FOR Q2 WITH R7 =a.2K
"20K,4W FOR 220V INPUT
FIGURE 11.9 TRANSISTORIZED 'ZERO VOLTAGE SWITCH
315
SCR·MANUAl
2) Transistor Qa supplies the triac triggering current (negative gate
drive) from the capacitor via Rli. Command signals to trigger
the triac can be inputed at the base of Qa.
3) Transistor Qb steering diodes Dl - D4 arid resistors Ri and
R2 constitute the zero voltage detector. When the line voltage
has risen 3-5 volts positively, current through diodes D 2 , D a,
R2 and Rs turn on Qb which in tum saturates Q2 and thereby
inhibits further gate drive. During the negative half cycle,
diodes Dl and D4 would conduct.
The triac triggering pulse will be about 100 microseconds long
for 115, 60 Hz operation and will be centered around the zero voltage
crossing point. Therefore the latching current available at the end of
this pulse determines the minimum load that can be successfully
controlled.
Operation of all these circuits (plus those to follow) on 50-400 Hz
is no problem, although a specially selected triac would be required
for 400 Hz supplies. All of the discrete elements respond fast enough'
so that the decision to trigger the power semiconductor is made before
the supply voltage exceeds 5 volts.
11.4 USE OF THE GEL300 - A MONOLITHIC ZERO
VOLTAGE SWITCH
The GEL300 is primarily a combination trigger circuit and
threshold detector to provide zero voltage switching control of resistance loads when used with thyristors such as triacs or SCR's. This
circuit provides a differential input stage, designed to sense resistance
bridges, with enough connections brought out to provide a wide variety
of useful connections. Output of this device can be adapted to provide
minimum RFI temperature control or to drive small relays or lamps.
Figure 11.10 shows the schematic of the GEL300. It is obvious
D,
D.
DO
D_
R.
••
K>
9 .
13
R_
aK
D.
".
R,
••
R.
.oK
eoQ
RO
aK
D.
••
FIGURE 11.10 CIRCUIT DIAGRAM OF THE GEl380 IC OFFERED IN A 14 PIN DIP.
NUMBERS REFER TO PIN CONNECTIONS'
316
ZERO VOLTAGE SWITCHING
that the main ditterences between this tigure and ...·igure 1l.~ is the
addition of an input stage consisting of transistors Qt and Q2 connected
in a differential amplifier configuration and a balanced resistor pair
(Rl and R2 ), which can be used as one side of a resistance bridge. The
circuit is so designed so that when Ql is conducting, its collector current
inhibits all output from the circuit (Q5 and Q6)' Otherwise, the IC
behaves exactly as explained before and the same precautions concefIling triac selection hold true (see also Chapter 12 for more details
concerning the use of the CEL300 in heating control circuits, in low
power circuits, staging heaters, sensors, etc.
11.4.1 Output and Power Connections
There is a wide variety of input and output connections for the
CEL300., The basic AC connection for triggering a triac is shown in
Figure 11.11. This figure is also the schematic Oess Rn and Rb), of the
power control (S200) module available from CE.
RA
8
Cs
+
13
10o,.F
15VOC
120 VAC
SO/60Hz
7
4
GEL300
RB
9
10
II
LU
3
(240VACI
10K
Rs
2W
(20KI
5W
RA 'THERMISTOR FOR TEMPERATURE CONTROL APPLICATIONS
FIGURE 11.11
BASIC POWER CONTROL CONNECTION
If it is necessary to control a pair of SCR's, the connection of
Figure 11.12(a) could be used. In this connection, the SCR's are driven
hy means of the pulse transformer T t • Since the CEL300 output pulse
is quite long, and normally starts before line zero crossing, it is necessary to shift the pulse so that the output of the pulse transformer occurs
at the proper time for triggering. This may best be accomplished by
advancing the pulse from the CEL300 with a leading network as
shown in the same figure. The output pulse to the SCR is taken from
the pulse that is generated from stored energy when the pulse in the
primary stops. This circuit scheme might also be used where isolation
between the line and the firing circuit is required. To provide a completely isolated low voltage control circuit, isolation transformer T 2 can
be added. T 2 need only supply about 15 rnA at 24 VAC.
317
SCR MANUAl
I.
AC
LINE
IK
3.3K
(a)
4o-1f----...
5
L~AD
FIGURE 11.12 SCR TRlcaERING CONNECTIONS
Figure 11.12(b) uses on SCR connected as a remote base symmetrical transistor, Qb to provide triggering for otber SCR's. Ql is a
low current SCR such as a C5 or a CI060f sufficient voltage rating for
the line voltage used. The output current from the GEL300 provides the negative base drive required by Ql for both half cycles. Gate
current for each SCR Hows from either D2 or D s, through Rl and Ql
to the gate of the required SCR. Diode D) is to protect the GEL300
in the event of Ql firing as a normal SCR due to a transient.
Since the output of the IC is limited plus the fact that the current
gain of the symmetrical transistor is so low, other means must be found
to trigger large SCR's.
318
ZERO VOLTAGE SWITCHING
11.4.2 Mating the IC to the High Current SCR
The circuits shown in Figure 11.13 show various ways of amplifying the IC output to a high enough level to guarantee a positive
tum-on of most any high current SCR. Since the GEL300 circuit was
designed specifically to trigger triacs, it is desirable to use a triac as a
pilot SCR wherever circuit voltage will allow its use, as shown in Figure
11.13(c). Figure 11.13(a) shows a useful technique for firing a pilot
SCR from a negative current source as is available from the GEL300.
The diodes Dl and D2 are needed to prevent excessive voltages from
being developed across the IC when initially turning on SCR 1 and
SCR2. Figure 11.13(b) provides positive slave firing of SCR 1 by means
of the PUT. R5 and C 2 may be adjusted to have the charge stored in
capacitor C 1 dumped into the gate of SCR 1 anytime before, during or
immediately after the cessation of current in SCR 2 This circuit does
away with many of the disadvantages associated with slave firing circuits having only passive components.
0
RI
SCRI
01
SCRa
°1
R3
47
r""-------,
+-....;-1- 0
I
I
°1
1
7
4
GEL300
1
I
:
FIGURE 11.13
4
I
I
GEUOO
:
7
1 7
I
I
~
I
(b)
1
4
GEL300
I
I
IL _____ ...II
L _______ ...II
IL _________ J
(a)
r----
r----- --,
1
I
(e)
HIGH CURRENT SCR GATE CIRCUITS TRIGGERED FROM THE GEl300
11.4.3 Connections for the Input Section
The basic connection of the input section as shown in Figure 11.14
is the direct comparison of a resistance bridge, using the internal resistors as one side of the bridge. RA and RR should be no lower than 5000
ohms in value to prevent undue loading when using the internally generated supply. The highest value of RA and RB may be determined by
the inhibit current (5 pA max) and the allowable error in the application. Obviously the next step could be to use both sides of the differential stage in an external bridge or to compare two external DC levels.
For temperature control applications, RA is usually a negative temperature coefficient thermistor.
319
SCH MANUAL
7
RA
8
INHIBIT
13
Ra
R2
8K
~
FIGURE 11.14 BASIC BRIDGE CDNNECTION FOR THE DIFFERENTIAL INPUT STAGE
The collector of Q2 (pin 8) can be used to generate a signal which
indicates the state of the input stage. H Q2is on, and drawing collector
current, then Ql is off and the circuit is supplying .output. This collector current can be sampled by a resistor (from 2 to 10 K ohm) to give
a voltage signal.
Chapter 12 includes more design ideas such as proportional control, staging of heaters to minimize light dimming, multiple triac triggering and others.
11.5 ZERO VOLTAGE SWITCHING AT HIGH FREQUENCIES
Many of the disadvantages of zero voltage switching at 50/60 Hz
disappear as the power frequency increases. These disadvantages
include:
1) Temperature excursions on small load heating systems. Since
the smallest increment of energy that can be applied is 1 or 1h
of a cycle, the temperature excursion may be excessive for
heater loads with small thermal mass. Thermal mass could
decrease inversely with frequency because energy absorbed is
proportional to time.
2) Lamp dimming. Zero voltage switching control of lamp intensity is not possible because of excessive flicker at low light
levels. Assuming that the eye is sensitive to any. event that
occurs less than 16 times a second, then at a 60 Hz, lamp
intensity can only be lowered to 1f4 brightness (1 pulse out of
every 4). However at 400 Hz, the lamp could easily bedimmed
to %0 of its full intensity (1 pulse out of every 20). Therefore
the solid state high frequency lrunp dimmer combines all the
advantages of solid state with very low EMI operation, especially critical on airplanes.
320
ZERO VOLTAGE SWITCHING
3) Motor speed control. Zero voltage switching of motors suffers
the same disadvantages of (1) above, namely excessive bursts
of speed for low jnertia loads. Therefore, one would except the
same improvement here at high frequencies as in (1) above.
11.& THREE PHASE ZERO VOLTAGE SWITCHING POWER CONTROL
Figure 11.15 shows the GEL300 inside the phases of a three phase
delta connected load. This circuit is a straightforward extension of the
GEL 300 used in single phase circuits of Section 11.4.1. It is applicable
to resistive or reactive loads by means of the circuit modifications
shown. Control maybe accomplished in two distinctly different ways.
The first method would involve three separate thermistors as shown,
each controlling one-third of the load. A separate type of control could
be accomplished by the use of a central control technique shown in
Figure 11.17 discussed later, which would synchronize all three load
legs with one central sensor.
NOTE: CONNECTED FOR RESISTIVE· LOAD.
FOR INDUCTIVE LOAD.
ADD DOTTED CXINNECTIONS & BREAK
X CXIM\IECTION.
--------------------+-----~
SCRI, 2 RATED fOR LOAD CURRENT a SUPPLY VOLTAGE.
240V. 3 PHASE SUPPLY60H z
ClRCUfT 8-C
a C-A ARE IDENTICAL
TO DETAl.. FOR A-B.
FIIIURE 11.15 THREE PHASE ZERO VOLTAGE SWITCHING CIRCUIT CONTROLLING PHASE
CURRENTS -OF DELTA CONNECTED LOAD
The power control circuit of Figure 11.16 shows the use of the
GEL300 controlling line current of a delta or wye connected load. R"
is used to establish an artificial neutral in conjunction with C 2 which
stabilizes the neutral such that it departs from zero voltage by less
than 4 volts p.p. on a 240 volt line. C 2 also provides a slight phase lag
to the GEL300 which guarantees that its narrow output pulse will
properly trigger SCR I .
321
SCR MANUAL
~~~~------------------------~'r~~-'r-----------------~
L _____ J
NaTES:
L CONNECTED FOR RESISTIVE LOADS.
FOR INDUCTIVE LOADS. AOIU)OTTED.CONNECTIONS
AlII DELETE COMPONENTs BRACl-f--<~~I-o;
R7
ISK
I
BANO· 2°r
PERIOD: 30 SEC.
5W
HE~rgR
6KW
MAX.
SET-POINT
ADJUST
usoon@70°F, loon/oF)
FIGURE 12.17
I
240V
60Hz
CALIBRATE
PROPORTIONAL CONTROL TEMPERATURE REGULATOR
Both of these circuits are compatible with the slaving and staging
circuitry, which will be discussed later.
338
SOLID STATE TEMPERATURE & AIR CONDITIONING CONTROL
12.5.4 Low Power Zero Voltage Switching Using the GEL300
When the GEL300 is used with a triac as a zero voltage power
switch, the triac latching current is the limiting parameter. If the current through the triac does not exceed the latching current at the end
of the gate pulse, the triac will return to the non-conducting state and
the control will therefore not operate correctly. For this reason, the
GEL300 specifications include a listing of special triacs which are
tested for latching at a minimum load current of 4.2 amperes.
To allow load less than the 4.2 amperes possible with the specified triacs, the gate pulse must be shifted in time. This allows the
conduction current to reach higher levels than would have been possible without the shift. A problem occurs if the pulse is shifted too far
and the gate pulse does not occur until after the line voltage zero
crossing. When this happens RFI is introduced into· the system. To
eliminate this RFI the pulse width must be extended.
The diagram and chart show how the gating pulse can be modified and the minimum loads which can be handled with each circuit.
120 VOLT SUPPLY
LlIlId
(watts)
Rl
00
500
00
400
00
250
200
6.SK
II.
R.
10K
7.5 K
7.5 K
4.7 K
0
2.2 K
2.2 K
2.2 K
C
0
0.01 /Ltd, 200V
0.022 /Ltd, 200 V
0.047 /Ltd, 200 V
240 VOLT SUPPLY
00
1000
00
SOD
500
00
400
6.S K, 1/2 W
20 K
15 K
15 K
6.S K
0
4.7K
4.7 K
3.3 K
0
0.0047 /Ltd, 400 V
0.01 /Ltd, 400 V
0.33 /Ltd, 400 V
Approx. Min. ApprDx. Pulse
Pulse
t/J Shift
100/Lsec.
100/Lsec.
100/Lsec.
150/Lsec.
0
25/Lsec.
50 /Lsec.
75/Lsec.
100,.sec.
100 /Lsec.
100/Lsec.
150/Lsec.
0
25/Lsec.
50/Lsec.
75/Lsec.
FIGURE 12.18 DESIGN CHART FOR LOW POWER LOADS
12.5.5 How to Use Low Resistance Sensors
The above circuits operate extremely well with negative temperature coefficient thermistors in the impedance range of 1 K to 10 K ohms;
but are unsuitable for very low impedance sensors. Most low impedance
sensors are used at the higher temperatures and are normally of the
positive temperature coefficient variety such as Nicrome, * tungsten
and platinum. For this sensors the circuit of Figure 12.19 has been
developed.
*Trademark of Driver-Harris Company
339
SCR MANUAL
LI
RI
33K
R3
330K
2N6027
Cs
IOo,.F
L2
*RSENSOR-WIND ENOUGH N, ORW WIRE TO EQUAL RSET (~IOD.f
NOTES:
I. ADJUST RI3 TO MID POINT BETWEEN ON AND OFF WITH C2 SHORTED, OSCILLATOR
DISABLED.
2. PROPORTIONAL CONTROL BAND (GAIN) DETERMINED BY Rg.
3. :~W r:~~;;E~~~~.N. PROPORTIONAL BAND IS 1% RSENSOR AND STROBE
4. A·PULSE T.RANSFORMER CONNECTED. BETWEEN (VI AND (Zl GIVES A SENSOR
.. ISOLATED FROM THE LINE.
'FIGURE··12.19
USE OF THE GEL300 WITH A lOW RESISTANCE SENSOR
The circuit includes a 2N6027 PUT relaocation oscillator which
operates at 20 pulses per second. When the PUT turns on it pulses a
resistor bridge consisting· of a reference divider (R4' R5 ) and the input
divider (Rseb Rsensor)' The dividers are coupled to the GEL300 inputs
by means of a capacitor and diode combination, such that with each
pulse, some charge on the capacitor is removed. The amount of charge
is proportional to the divider ratio. The resistors R s, R7 and Ro serve
to reset the coupling capacitors between the pulses. Since each divider
is connected to opposite sides of the input differential amplifier of the
GEL300, the GEL300 will tum on if the R.pt, R."nHOr divider is unbalanced such that point Y is higher than the 50% point of the voltage
between X and Z. For positive temperature coefficient sensors this i"
equivalent to an under temperature condition.
An interesting modification of this circuit involves isolating the
sensor from the line. To accomplish isolation of the sensor, one needs
only to connect a pulse transformer (such as a Sprague llZ12 or a
Pulse Engineering PE-2229) between points Y and Z. The low resistance sensor is then connected to the pulse transformer secondary to
complete the circuit.
12;5.6 .Multiple Triac Triggering
If more than one triac is required additional triacs can be added
in the manner shown in Figure 12.20. In this circuit, the GEL304
Threshold Detector serves as a buffer amplifier between the GEL300
and the .triac gates. With 50 rna I gt triacs, five triacs can be driven
from each GEL304. There are other advantages of this' approach. The
first is that by increasing Rand C so that the time constant is about
10 ms, integral cycle control can be obtained. This ensures that inductive loads are switched for full cycles eliminating the possibility of
saturation.
340
SOLID STATE TEMPERATURE & AIR CONDITIONING CONTROL
~4r~--~----~----~--~~~--~------------~--~-
•
TO FOLLOWING
STAGES AS
REQUIRED
6811
EA.
120Y
(240Y)
TO OTHER
GEL304'S
AS REQUIRED
TO OTHER
TRIACS
AS
REQUIRED
10K
(2Ok)
INTEGRAL CYCLE CONTROL CAN BE OBTAINED BY CHANGING R TO 1.2 NEG. AND C TO O.OI,..F
FIGURE 12.20 IIEL300/GEL304 TRIGIIER CIRCUIT FOR DRIVINII MANY TRIACS
The second advantage is that with the widening of the pulse width
triacs no longer need be selected for latching and pulse gate trigger
current. Pulse widths of 200 p.Sec will guarantee that all triacs will be
latched on by that time and the gate trigger current can be considered
de for such a wide pulse.
12.5.7 Load Staging
Many times when multiple loads are to be controlled, it is desirable to sequentially energize them. Figure 12.21 shows a method of
doing this. The circuit is an extension of Figure 12.20, with the RC
time constant set greater than the period of the 60 Hz line voltage so
that if the GEL300 calls for load power the adjacent GEL304 will be
in the on condition for at least the next full cycle.
I
r~MPERATURE
z~J.:~~grc
LINE
NEUTRAL
FIRST
..
iR~~~~
I
..
SECONDH
STAGE
DRIVER
It:
~1GD
It:
III
+
l::
10Op.F ;: :::-1....
lOOK
< IK
onli!
7
S
~ GEL300FI 14
2
5
OlSO!;
4
OZSO!;
I
4
GEL304AI 13
2
10K
I ~ GEL304AI
P.
2
5K
O.8p.F;:::::;:
1.0p.F;:::r:
10K
2W
68D.
6Sa
TO
OTHER
GATES
-9V
TO
OTHER
GATES
LINE
120V
w
,...
FIGURE 12.21
fii:\
'-lYr.f
\!Y
STAGED TEMPERATURE CONTROL
341
SCR
MANUAL
When this GEL304 is on it begins to discharge the 1.0 pF capacitor coupled to its output through the 10 megohm resistor. After
6 seconds the second GEL304 will turn on and trigger the triacs connected to its output. If the GEL300 output stops pulsing the first
GEL304 will turn off resetting the 1.0 pF capacitor quickly through
the bypass diode around the 10 megohm resistor. This then restarts
the timing delay. By varying the 1 p.F capacitor or the 10 megohm
resistor delay times from one cycle to several minutes can be obtained.
If this staging mechanism is used with a proportional control system,
such as the one described below, one could set the delay time greater
than the repetition rate of the proportioning circuit. Under this condition the staged loads would only be energized when the system is
operating outside the proportional band and full power is being
required.
12.5.8 Fail.Safe Operation
Positive temperature coefficient sensors have an inherent advantage - fail safe operation. If a P.T.C. sensor is broken or opens, it
appears as if an over-temperature condition exists and no power is
delivered to the load. Negative temperature coefficient sensors lack this
advantage; in some applications this aspect should be considered in
the design. To avoid this possibility, the circuit of Figure 12.22 may
be employed. In this circuit, if the sensor opens, the PUT turns on
providing base drive to the transistor. The transistor then shorts the
supply capacitor of the GEL300 rendering it inoperative. When the
capacitor is discharged the PUT will turn off but will turn back on
when the internal supply of the GEL300 attempts to recharge the
capacitor. When the sensor is replaced the system will resume normal
operation.
+
GEL~OFI
2
FIGURE 12.22
SENSOR OPEN DETECTOR FOR GEL300F1
.12.6 AIR CONDITIONING
Up to this point, the discussion has mostly entailed electric heating with some small mention of phase control to regulate motors. However, heating is but one facet of space conditioning. Other aspects are
cooling, ventilating, and heat distribution systems. In all (neglecting
thermoelectric cooling), motor control plays a key role. While Chapter 10 is devoted exclusively to this subject, specialized circuits have
been developed and tested (1) for these roles and should be mentioned.
342
SOLID STATE TEMPERATURE & AIR CONDITIONING CONTROL
12.6.1 Cooling
Most systems, with the exception of hot water, utilize a condenser
and compressor to cool room air. A majority of these systems allow
these cooling components to have hysteresis, i.e., the temperature at
tum-off is lower than that at turn-on, reducing the number of times
the cooling components operate and thus increasing their life. The
mechanical method is to utilize a mechanical thermostat to turn the
units on and off directly or, for the larger units, through an electromechanical relay. In a solid state system, considering the relatively high
compressor surge current for the larger units (60 amps or greater), and
the desire to isolate the thermostat from the line voltage, it is presently
economically advisable to retain the electromechanical relay. To this
end, the circuit of Figure 12.23 is a solid state replacement for a
mechanical thermostat designed to control a condenser fan and
compressor.
•
RI
RI2
.
:;
0
>
0
;:;
.....2
0
OB
'".
0:
~
z
l!:
Z
0
u
@
.
::.'"
R.
0:
0
0
u
®L---~~--w..___-..I
~
COMPRESSOR
OFF
Rio R4. R5. R. oRu. RI3 - IN 6EL500
Rt.R!. RIB
- IK ohm
R6
RT
fig!
RIO
RJ2
- 5 megohm
- 12k otIm
-~ohm
- 5Kohm
- 2.2K ohm
RI4
RI5
- e2K ohm
- 220Koom
RT
- 6E 2RII4
R
-120ohm
- G£ 43F9723AA9 25o".F 25V
- Gf 75FIR5AI03 .o1~F SOV
- IN GEL300
D.~.D3,D4 - GE AI4F
01 thru 06 -IN GEL30D
C
CI
DI
07
- 2N5354
08
Z
- GE CI06F
- Gf 14XLI2
KI
- 24V COIL, CONTACTS AS REO.
220V PRI
{ 24V CT. SEC.
2' WATT
11
NOTE:
ALL RESISTORS 112W 1:.10% UNLESS OTHERWISE NOTED.
FIGURE 12.23
COOLING COMPRESSOR AND CONDENSER CONTROL
The condenser fan and compressor are energized by Qs through
relay K1. Qs is turned on Simultaneously with Q7. Consequently it
remains to turn on Q7 whenever room cooling is required.
The GEL300, used here as a level detector with hysteresis, is utilized to control Q7 and consequently the condenser fan and compressor.
Note that the GEL300, shown in thin line, is used in the DC mode
only and does not utilize the AC synchronization components internally
connected to its pin 5. That is, Q5 and Q6 will be off and Q7 on through
R 12 , D 2, D a, Qs and relay Kb regardless of the AC power signal, if Ql
is on as dictated by the thermostat components. Therefore, when the
room temperature is high, Ql is on, Q5 and Q6 are off, Q7 is on, and
power is applied to the condenser fan and compressor. At this time
343
SCR MANUAL
R6 , R I5 and RI2 are across Rb R2 and part of R3. As the room temperature decreases, the increasing resistance of RT will overcome QI and
tum on Q2, Qo and Q6, removing power from the cooling components,
allowing room temperature to increase. Note that R6 , R I5 and R I3 are
now across R4 and part of R3. The changing position of R6 and RIo
introduces hysteresis and requires the room temperature to increase
somewhat before the resistance of RT reduces sufficiently to allow Ql
to conduct and ultimately re-energize the cooling components. The
hysteresis of this control is adjustable from approximately O.25°F to
4"F by R6 •
R7 isolates the thermostat from Q2 interference and C 1 eliminates
erratic relay closure due to noise. The circuit was designed to be operated with mechanical overload protection to prevent rapid, continuous
compressor current surges. However, if desirable, this feature could be
accomplished electronically.
12.6.2 Ventilating
A proportional motor control which varies the speed of a ventilating fan or blower in response to a heating or cooling requirement, aids
in reducing room temperature variations, drafts, noise and variation in
noise. (3) These advantages have been well known in the room conditioning industry but were not economically feasible until the arrival
of solid state. Since a ventilating motor does not usually require the
high current capability of inverse parallel SCR's, a triac power switching component is generally used. The triggering circuits may be simple
or sophisticated, depending upon the accuracy desired. The thermistor
thermostat is similar to that used for heating or cooling, however, it
may not be located within the room, e.g., it may be located within a
heated or cooled chamber.
In the customary electromechanical furnace central heating system (oil, gas, etc.), the room thermostat controls the action of the burner
and a second thermostat, mounted in the bonnet, controls the action
of the blower. When the room thermostat calls for heat, the burner is
energized and bonnet temperature begins to rise. When bonnet temperature reaches a predetermined high-temperature limit, the blower
is energized to circulate the heated air. When the room thermostat is
satisfied and de-energizes the burner, the blower continues to run until
the bonnet temperature drops below a given low temperature limit,
at which point the blower is turned off. This on-off cyclic action results
in room temperature variations that are beyond the control capability
of the room thermostat. (4)
'
Figure 12.24 shows Generr..l Electric's SlOOE solid state speed
control assembly for the furnace blower that replaces the bonnet thermostat with a thermistor. This assembly provides continuous control
of blower speed iIi response to bonnet temperature. It also limits the
minimum speed at which the motor can run which protects the bear344
SOLID STATE TEMPERATURE & AIR CONDITIONING CONTROL
BLACK
RED
(OJ
FIGURE 12.24 FURNACE BLOWER MOTOR SPEED CONTROL
ings, and maintains a gentle circulation of air through the heating
system. When the room thermostat energizes the burner, the blower·
speed will increase gradually as bonnet temperature increases. Heat is
thereby distributed to the house as soon as it is available from the
burner and the thermostat then has the opportunity to tum off the
burner long before the full capacity of the system is reached. The effect
greatly reduces the temperature excursions that can be experienced
in mild weather. Under these mild heating load conditions, the blower
may never reach full speed in order to maintain, the proper temperature of the house. Conversely, in severe weather, the blower may never
reach minimum speed because of the high demand for heat. The control is capable of controlling up to 10 amperes -on a 230 volt line. It is
contained in approximately a I" x 2" x 3" metal housing equipped with
pigtail leads, as shown in Figure 12.24, to which are connected the
user's supply voltage, motor load, thermistor Tand potentiometer R l •
The detailed operation of the control is as follows. When the triac
is pulsed on through either diac, it applies power to the motor load until
the current reduces to near zero at the end- ·of its half cycle. The triac
then turns off, removing power from the motor, allowing the triggering
circuits, consisting of C b C 2 , Rb Ra and the two diacs, to start timing
again with reference to the voltage across the triac. As mentioned
earlier, this type of trigger synchronization may result in an unsymmetrical AC waveform applied to the motor, but for the intended
S100E applications it is generally satisfactory. When the bonnet temperature is low, the resistance of Ra will be high, and the time required
for C h charging through Ra, to reach the breakover potential of its
diac will be long. Therefore, depending upon the setting of Rh it is
possible that C 2 will reach the breakover potential of its diac first,
guaranteeing the minimum speed limit. As the temperature of the bonnet increases, with resulting decrease in Ra, C 1 will fire the triac
through its diac earlier in the cycle than C 2 and R!J are capable -of,
increasing motor speed.
345
SCR MANUAL
C a and R2 prevent dv/dt triggering and, assisted by Ll and C 4 ,
smooth the steep voltage wavefront, reducing RFI to a tolerable level.
12.6.3 Ventilating Blower Control for Heating and Cooling
Some room conditioner systems are designed with. a ventilating
blower (or fan) proportionally controlled during the heating and cooling cycles. This has proven to provide excellent room 'temperature
regulation.
.
.
The circuit of Figure 12.25 is: such a control. It is designed to
be operated from one thermostat located within the room. When
neither room heating nor cooling is required, the control operates
the blower at minimum speed. As room temperature decreases from
SP, blower speed is proportionally increased. When the heating source
increases room temperature the control will proportionally reduce
the blower speed. Likewise when room temperature increases from SP,
blower speed is proportionally increased and then proportionally
decreased when the cooling source lowers the room temperature. From
the previous discussion it should be noted that the control is capable
of proportionally increasing blower speed for either an over or under
temperature deviation from SP. In addition it has RFI and dv/dt suppression, line voltage synchronization with a continuous triac gate signal for good motor performance, minimum speed limit, thermostat
isolation and is capable of controlling 6 amperes at 240 volts. Higher
currents are possible, but probably not needed, by utilizing a larger
power triac.
R , • RZI ' R 23 - II< ohm
RZ' R4
- 4.11< ohm
"3
- 2.51< Ohm
"."."7
- lOOK ohm
Re,R,g
".
"'a
"
Rll , R'8
A13 , R I5
",.
",.
- 50K ohm
"11
- I.SK ohm
- 15K ohm
- 3.31< ohm
R24.RZ5
R
- 5k ohm
- 4701< ohm
"20."22
RT
C
- GE43F9T23AA9 250J.'f 25V
09 -GESCI41D
- 2.2K ohm
C,
- GE75FIR5A4T2 .0047#4' SOV
Z
- 680 ohm
Cz - GEAAI4AI04A .IILF IDOV
T2
- 10K ohm
C3 - GE75F7R4-224 .22p.F 400V
- 220K ohm
- 3.91< ohm
- 47 ohm
C4 - GETSF4R4-473 .047,&.1.' 40aV
P"'
T, {220V
24V CT SEC
- 33K ohm
- 120 ohm
- GE2RII4
D. D, thru D4 - GEAI4F
0,
- 2N5354
L,
-GEZ4XLl2
THORDARSON. 23Vl23
25 WATT
- 100ft"
Q2 - GE3N86
Q3 thru. Q7 - GE2N3393
Os - GECI06Y
NOTE:
ALl RESISTORS 112 WATT .:tIO'%. UNLESS OTHERWISE NOTED.
FIGURE 12.25 VENTILATING BLOWER CONTROL (HEATING AND COOLING)
346
SOLID STATE TEMPERATURE & AIR CONDITIONING CONTROL
The control is capable of being operated in conjunction with the
cooling compressor and condenser control of Figure 12.23, therefore
it utilizes their common components of TJ, D, C, R, Z and the thermostat, consisting of RD, RlO and thermistor R T •
The detailed operation is as follows:
When the triac Q9 is triggered through T 2 by the triggering circuit, which consists of those components connected to the low voltage
side of transformer T], it applies power to the ventilating motor.
T2 performs three functions: (a) provides coupling from the
isolated triggering circuit to the triac gate; (b) supplies a continuous
gate signal prior to triac tum-on; (c) automatically provides a symmetrical negative half cycle triac gate signal after once being shorted
during the positive half cycle by Qfl. Functions (b) and (c) require some
explanation. Note that the primary (220 volt side) of T 2 is connected
to the gate of the triac. With its secondary open circuited, and the
triac off, its impedance is high and, combined with the shunting effect
of R24 , will not allow the triac to tum on. Shorting the secondary of T 2
some time during the positive half cycle causes a large primary current
to flow into the gate of the triac. The current continues until the triac
turns on, shorting out the primary of T:!, stopping gate current and not
allowing the transformer to change its flux level or direction. Removal
of the short at the start of the negative half cycle and ultimately the
turning off of the triac, results in the transformer preventing gate current until the supply voltage causes it to saturate during that negative
half cycle. The transformer then permits sufficient tum on current until
the triac turns on, again shorting out the primary of T:!, stopping gate
current and not allowing the transformer to change its flux level or
direction. Therefore, at the end of the negative half cycle the transformer core is reset ready to be shorted again during the positive half
cycle. After a few cycles, the phase angle of negative cycle saturation
will be very close to the phase angle at which shorting took place during the positive half cycle. Consequently T:! provides a continuous gate
signal during the negative half cycle until the triac turns on, symmetrical with that of the positive half cycle. The symmetry is best with a
square loop core. However, the transformer used is a standard filament
type and results in very good motor performance. It should be noted
that Qs, triggered by Q:!, shorts T:! during the positive half cycle and
at the desired phase angle.
Synchronization to the voltage supply is accomplished by Q •.
During the positive half cycle Ql is saturated and the DC supply voltage is applied to the triggering circuit. During the negative half cycle
Ql is cut off, removing the DC potential, allowing T:! to provide the
predetermined negative half cycle symmetrical trigger.
Q:! is turned on when its anode potential is approximately 0.6 volt
higher than its gate, as determined by the ramp and pedestal charging of capacitor C=.!. That is, C:! charges very rapidly through Q:\ or Q4
and R7 and then continues to slowly charge through R" and R/i, providing a high gain, well defined trigger for Q:!. R/i is then used as a reasonably independent minimum speed adjustment.
347
SCR MANUAl
Qa and Q4 allow the control to increase blower speed for either
an over or under temperature deviation from SP. At SP, R21 is adjusted
so that the potential at the emitters of Qa and Q4 are equal. At this
timeC2 is charged at a rate which results in the blower rotating at its
minimum speed. When the temperature decreases, Q7 decreases conduction and the voltage at. Qa rises, charging· C 2 faster, increasing
blower speed. When the temperature increases, Q6 decreases conduction and the voltage at Q4 rises, again charging C 2 faster and increasing
blower speed. The rate at which the blower increases speed with
respect to temperature deviation is very dependent upon th gain of the
Q6 Q1 differential amplifier as dictated by the R n , Rg, R17, RID resistanCf' ratios.
Capacitor C 1 reduces noise interference. Q5 reduces the interference into the thermostat. R25 and C s prevent dv/dt triggering and,
assisted by LI and C 4, smooth the steep voltage wavefront, reducing
RFI to a tolerable level.
12.6.4 Fan and Coil Blower Controt
Figure 12;26 shows a proportional motor speed control especially
designed for fan and coil water systems. A fan and coil water system
is a room temperature regulator capable of heating or cooling by means
of passing hot or cold water, from a central supply, through heat
exchanger ·coils in each room wh, ore a blower helps transfer the heat
from or to the room air. The purpose of this solid state control is to
improve room temperature regulation by proportionately controlling
blower speed in accordance to room temperature demands as indicated
by the room thermostat.
Since heating or cooling is accomplished by passing hot 01' cold
water respectively through one coil, the control must again be capable,
as in Figure 12.25, of increasing blower speed for either an over or
under temperature deviation fromSP. In addition it has RFI and dv/dt
suppression, a minimum speed limit and is capable of controlling
6 amperes at 110 volts. Higher currents and voltages are possible but·
probably not necessary. The triggering circuit is shown within the
dotted block and is synchronized to the power triac, Qt. This type of
synchronization is acceptable for the majority of the motors used in
this application.
Ql is triggered on with a pulse supplied by unijunction Q", through
T 1. RlI and C 4 determine the minimum speed. When a unijunction
firing signal is supplied through D7 later in the cycle than that supplied
through Ds as a result of C 4 charging through R H , the blower rotates
at the minimum speed dictated by Ru. When the central water control
determines that cooling is required, and provides water colder than
normal room temperature, the water sensing thermistor T w causes Q2
to be off and Qa on . .This requires C a to charge through R4 and the room
temperature thermistor T A. If the relative resistance of TA with respect
to the SP potentiometer R5 is such so as to request room cooling, C a
will charge in a ramp and pedestal fashion(2) to the unijunction firing
voltage ahead of C 4, increasing blower speed and decreasing room
348
SOLID STATE TEMPERATURE & AIR CONDITIONING CONTROL
r------- - - - -- - -- - ----------,
MI
r----'
t;\
I
I
I \:}
I
I
I
R2
L ____J
C2 RI
Q4
~------------------------R,
R2
- S2 ohm
·4.7K.4W
%. R'2
-IK
R4. R5. R7 - 5K POT. 112W
R6.R9
-3.3K
RS
-22K
R,O
- 4.7<
-IOOohm. II~W
RII
C, -.221£. 2vvV
C2 -.05JL. ~OOV
C3 -.1".50V
C4 -.1".5CN
0, Ihru 04 - GE 6E8102
~ Ihru De - AI4F
Q, - TRIAC A5 REQUIRED
Q2. Q3 - 2N2712
q4 -6E 2N2646
z, - Gr ?4XL.:::v
Tw - ¥~~~~~ST~"'.
TA -
25°C
~~~~~~siokR Q 2~"C
M, - 3 AMP SHADED POLE MOTOR
T, -
~~~~~~~~2/ULSE
NOTE:
ALL RESISTORS 112W tlO'¥. UNLESS OTHERWISE SPECIF lED.
FIGURE 12.26
FAN AND COIL BLOWER SPEED CONTROL. TEMPERATURE REGULATOR
temperature. As more cooling is required the blower speed will be
modulated up to its maximum. In the event the central water control
determines that heating is required and provides water warmer than
nominal room temperature, T w forces Qa off and Q2 on. Now Cacharges
through R 7 , R6 and R 5. Consequently when the relative resistance of
T A with respect to R5 is such so as to request room heating, C a will
again charge to the unijunction firing voltage ahead of C 4 , increasing
blower speed and rooni temperature. Thus blower speed is increased
for either an over or under temperature deviation from SP.
Rl and C 1 prevent dv/dt triggering and, assisted by Ll and C 2 ,
smooth the steep voltage wavefront, reducing RFI to a tolerable level.
REFERENCES
I. Penkalski, T. A., et aI, "Optimum Solid State Control Parameters
for Improved Performance of In-Space Electric Heating Systems,"
General Electric Publication 671.12.
2. Cape, R C. and Tull, R H., "Test Room Performance of LineVoltage Thermostats," IEEE Conference Record of 1967 Industrial
and Commercial Power Systems and Electric Space Heating and
Air Conditioning Joint Technical Conference, May 22-25, 1967.
3. Application Data 3702, General Electric Company, Edmore,
Michigan.
4. Howell, E. Keith, "Switch From Hot to Cool," Electronic Design,
February 15, 1967.
349
SCR MANUAL
NOTES
350
CHOPPERS, INVERTERS AND CYClOCONVERTERS
13
CHOPPERS, INVERTERS AND
CYCLOCONVERTERS
This chapter describes choppers, inverters and cycloconverters
using SCR's which perfonn the functions previously performed by
electrical machines, mechanical contacts, spark gaps, vacuum tubes,
thyratrons and power transistors. These functions include standby
power supplies, vibrator power supplies, radio transmitters, sonar transmitters, variable-speed AC motor drives, battery-vehicle drives, ultrasonic generators, ignition systems, pulse-modulator switches, etc.
The advantages of using equipment with solid-state switches to
perform these functions are;
Low maintenance
Reliability
Long life
Small size
Light weight
Silent operation
Insensitivity to atmospheric cleanliness or pressure
Tolerance of freezing temperatures
Operable in any attitude
Instantaneous starting
High efficiency
Low cost
13.1 CLASSIFICATION OF INVERTER CIRCUITS
The following definitions are used in this chapter;
Rectifier:
Equipment for transforming AC to DC
Inverter:
Equipment for transforming DC to AC
Equipment for transforming AC to AC
Converter:
DC Converter:
Equipment for transfonning DC to DC
Cycloconverter: Equipment for transforming a higher
frequency AC to a lower frequency without
a DC link
Cycloinverter;
The combination of an inverter and a
cycloconverter
Chopper:
A "single ended" inverter for transforming
DC to DC or DC to AC
Note: The term inverter is also used in this chapter as a generic
term covering choppers, inverters, and the several forms of converters.
Thus "Classification of Inverters" covers classification of Choppers,
Inverters, Converters, and DC Converters.
351
SCR MANUAL
13.1.1 Classes of Inverter Circuits
The basic classification of inverter circuits is by methods- of turnoff. These have been described in Chapter 5. There are six classes:
Class A SeH commutated by resonating the load
Class B Self commutated by an LC circuit
Class C C or LC switched by a load-carrying SCR
Class D C or LC switched by an auxiliary SCR
Class E External pulse source for commutation
Class F AC line commutated
SCR INVERTERS
CLASSES
I
A
B
C
SELF COMMUTATEO
BY RESDNAnNG
THE LOAD
SELF COMIIUTIU"EO
BY AN LC CIRCUIT
C OR LC SWlTCHEO
BYLOAO-_e
SCR
123456
123456
CONFIGURATIONS
J
I
rJm In
I
D
E
F
LC swm:HED
BY AUXILIARY
SCR
EXTERNAL PULSE
SOURCE FOR
CO....UTATION
AC LINE
CO....UTATEO
!HI!!!
123456
rtrn
!;:
C
3=
4·
=
5
6= TH
I
I
rJm
I
~
123456
TfP'lA~:bION
TAPPED SUPPLY
PHASE HALF WAVE
PHASE FULL WAVE
TABLE 13.1
13.1.2 Properties of the Inverter Classes
Class A - SeH commutated by resonating the load. These inverters are
most suitable for high-frequency operation, i.e., above about 1000 cps,
because of the need for an LC resonant circuit which carries the full
load current. The current through the SCR is nearly sinusoidal and so
the initial dildt is relatively low. Class A inverters lend themselves to
output regulation by varying the frequency of a pulse of fixed width
(time ratio control).
Class B - SeH commutated by an LC circuit. The great merit of this
class is circuit simplicity, the Morgan chopper being an outstanding
example. Regulation is by time ratio control. Where saturable reactors
are used, some skill is necessary in the design of these components,
and manufacturing repeatability must be checked.
Class C - C or LC switched by a load-carrying SCR. An example of
this class of inverter is the well known McMurray-Bedford inverter.
With the aid of certain accessories this class is very useful at frequencies
below' about 1000 cps. External means must be used for regulation.
Class D - L or LC switched by an auxiliary SCR. This type of inverter
is very versatile as both time-ratio and pulse-width regulation is readily
incorporated. The commutation energy may readily be transferred to
the load and so high efficiencies are possible.
Class E - External pulse source for commutation. This type of commutation has been neglected. It is capable of. very high efficiency as
only enough energy is supplied from the external source for commutation. Both time-ratio and pulse-width regulation are easily incorporated.
Class F - AC line commutated. The use of this type of inversion is
limited to those applications where a large amount of alternating power
is already available. Efficiencies are very high.
352
CHOPPERS, INVERTERS AND CYCLOCONVERTERS
13.1.3 Inverter Configurations
Rectifier circuits occur in several configurations such as half-wave,
full-wave, bridge, etc. Inverter circuits may be grouped in an analogous
manner.
Figure 13.1 shows the different types of configurations. Methods
of triggering and commutation have been left out for clarity.
I. CHOPPER
3. CENTER TAPPED SUPPLY
2. CENTER-TAPPED LOAD
4. BRIDGE
5. THREE PHASE HALF WAVE
6. THREE PHASE BRIDGE
FIGURE 13.1
INVERTER CONFIGURATIONS
13.1.4 Properties of the Different Inverter Configurations
CONFIGURATION
1
2
3
4
5
6
Chopper
CT Load
CT Supply
Bridge
3 Half-Wave
3 Bridge
Blocking Volts(1)
E
2:E
E
E
E
E
Peak Load-Volts
E
E(2)
1/2 E
E
E
E(3)
yes
no
no
no
yes
no(4)
Numi1er of SCR's
1
2
2
4
3
6
Ripple Frequency
in Supply
f
2f
f
2f
3f
6f
1
1/2
1
1/2
1/3
1/3
yes
no
yes
yes
yes
yes
DC in Load
Ave SCR Current(1)
Supply Current
Transformer-less
Operation Possible
(l)
(2)
(3)
(4)
Ignoring overshoot due to commutation.
Using a 1:1:1 transformer.
Line-to-line voltage.
Assuming symmetrical loading.
353
SCR MANUAL
13.1.5 Discussion. of Classification System
This method of classification gives thirty-five (35) different classes
and configurations. However there are many circuits which could fall
into the same classification and which are yet different. This occurs
particularly in Class D where the method of commutating the auxiliary
SCR may take many forms. There must therefore be several hundred
possible inverter circuits.
In the following pages five examples are given to illustrate the
scope of SCR inverters and the design procedure. The examples cover
perhaps 1 % of the possible circuit variations. It is for the equipment
designer to use the classes and configurations together with the accessories to be described as building blocks to form the best combination
for this particular application.
13.2 TYPICAL INVERTER CIRCUITS
13.2.1 AClass AInverter
The design of Class A inverters has been well covered in the literature.! The following data, taken from Reference 1.4 of Section 13.6,
illustrates the performance of one Class A inverter. Space does not permit the inclusion of the development of design procedures.
13.2.1.1 Circuit Description
The operation of the circuit is as follows. In Figure 13.2 when
SCR I is triggered, current Hows from the supply El charging up capacitor C to a voltage approaching 2E I . The current then reverses and Hows
back to the supply via diode Dl and C discharges. During the reverse
current How, turn-off time is presented to SCR I . SCR:! is triggered next
and a similar cycle occurs in the lower half of the circuit with a negative going pulse of voltage appearing across C. SCR I is now triggered
again and so the cycles repeat.
01
39n
GE A280~
o.0221'F
=
-=- E\150V
3mH
LOAD
FIGURE 13.2 A CLASS A INVERTER CIRCUIT
354
CHOPPERS, INVERTERS AND CYCLOCONVERTERS
Figures 13.3(a) and (b) show the circuit waveforms with no load
and full load respectively. A comparison reveals some of the features
of this new inverter. The output voltage, the peak SCR voltage, and
the output voltage waveform remain virtually unchanged.
{al NO LOAD
{bl FULL RESISTIVE LOAD
{el FULL CAPACITIVE LOAD
{d 1 FULL INDUCTIVE LOAD
(A) SCR and Diode Current
(B) Output Voltage
(C) SCR Voltage (Anode to Cathode)
FIGURE 13.3
CLASS A INVERTER WAVEFORMS FOR CIRCUIT OF FIGURE 13.2
Figure 13.3(c) shows the effect of a heavy capacitive load on the
circuit waveform. Note the broadening of the current pulses and
the increase in output voltage. Figure 13.3(d) shows the effect of ,1
heavy inductive load with an opposite trend. Neither leading nor lagging zero power factor loads have any serious effects on turn-off time
or component voltages in tihs inverter circuit.
Figure 13.4 shows the effect of varying the triggering frequency
(fo) on the output voltage waveform while keeping the resonant frequency (fr) of the LC circuit constant. Lowest distortion is seen to
occur at a ratio of f.lf"
1.3.5.
=
355
SCR MANUAL
fr/fo'
2
• 1.5
-1.35
• 1.2
• 1.1
FISURE 13.4
EFFECT OF VARY INS RATIO OF RESONANT FREQUENCY OF LC
CIRCUIT f. TO TRIGSERING FREQUENCY f. ON OUTPUT
WAVEFORM OF CLASS A INVERTER (CIRCUIT OF FISURE 13.2)
The features of this inverter are:
1. Good output waveform.
2. Excellent load regulation.
3. Ability to operate into an open circuit.
4. Ability to work into a wide range of reactive loads.
5. Relatively low and constant value of SCR voltage.
Figure 13.5 gives the calculated and measured load regulation
curves for reactive and resistive loads of the design described in detail
in Reference 1.4 of Section 13.6.
II.
...
12.
-
.,.....J.,~ -1
• ---'
o·
I
RDISTNE
•
to- INDucTIVE
I
•-
I
...........
-
CALCULATED
-
4.
2.
,
•
·LOAD CURRENT IN
FIGURE 13.5
AMPS
10
CR_)
12
14
.
LOAD REGULATION OF CLASS A INVERTER OF FISURE 13.2
13.2.1.2 Applications
The following are some of the uses for Class A inverters:
• Ultrasonic cleaning, welding and mixing equipment
• Induction heaters
• Radio transmitters in the VLF band
• Sonar transmitters
• Cycloconverter supplies, the output of the cycloconverter
itself being useful for all applications where AC power is
used
356
CHOPPERS. INVERTERS AND CYClOCONVERTERS
• DC to DC converters where the advantages of light weight,
small size, low cost and fast response time due to the highfrequency link are very apparent
13.2.2 AClass BInverter
While many examples exist in the literature of Class B choppers,
until recently, Class B inverters were not widely discussed. Recent
literature has reported on novel developments utilizing the Class B
principle in DC to AC conversion. The following data, taken from
References 8.12 and 8.22 of Section 13.6 discusses the principle of
operation and pedormance features of one Class B inverter. Space does
not permit the inclusion of the development of design procedures.
Design equations and notes are included in the cited references.
13.2.2.1 Circuit Description
The Class B regulated sine wave inverter is derived from the
basic Class B tuned inverter circuit of Figure 13.6. The basic circuit
suffers from three major drawbacks which the circuit of Figure 13.8
overcomes. The circuit pedormance limitations are: instability and
severe transients upon load disruption due to stored reactive power,
lack of means for voltage regulation and sensitivity to load power
factor. In spite of its sinusoidal output characteristics these disadvantages have severely limited application of the basic circuit.
J
= LI
t:"::\
P
-=- DC SOURCE
t
;: ::;:CI
L
0
A
'1
D
f.::\
I
TRIGGER
CIRCUIT
FIGURE 13.6
BAS IC CLASS B TUNED INVERTER
357
SCRMANUAL
Figure 13.7 shows the first of two major modifications introduced
to the basic circuit. The purpose of the additional components, L 2 , La
and D t through D 4, is to provide a path for rapid return of excess
reactive current. This passive stabilization network eliminates unwieldy
voltage transients and stabilizes the output voltage for large abrupt
changes in output loading. Its main feature is its ability to accomplish
the above goals without substantial clipping or distortion of the output
wave shape. This is accomplished by returning the feedback current
through a differential choke L2 such that equal current is distributed
between Points A and B of Figure 13.7. Point B contains a considerable
amount of pulsating voltage which would normally severely affect the
feedback current and consequently cause output voltage distortion. By
use of L2 both the clipping that would be present using solely Point A
for a return path, and the distortion of using Point B is eliminated.
Diodes D t and D2 serve to isolate Ll from L 2 • La is used to limit
extremes of feedback current while being dimensioned small enough
to provide a short time constant for good short time feedback performance under transient conditions. Diodes Da and D4 serve to rectify the
feedback current for return to the supply. It has recently shown that it
is possible to combine L] and L2 by judicious tapping of Lt. See Reference 8.12 of Section 13.6.
_L_I_
AJ·~18
1
DI *
L2
SCRI
TI
D3
*D21
o-/'V'"" " " ' - -
IVVYl
fDCSOURCE
;:: r:CI
1
~
D4
L3
~
--
L
0
A
D
SCR2
r
TRIGGER
CIRCUIT
FlGURE'13.7
SINE WAVE INVERTER WITH PASSIVE FEEDBACK STABILIZATION
Figure 13.8 shows the completed circuit with the added second
addition of an active feedback system capable of providing voltage
regulation for input source and load variations. The active regulating
circuit is comprised of C 2, L 4, and SCRa and SCR 4 • Operation of the
active feedback circuit is controlled by SCR's 3 and 4. By varying their
phase angle triggering, the voltage on capacitor C 2 can be made to
vary, such as to either add or subtract from the source supply voltage.
358
CHOPPERS, INVERTERS AND CYCLOCONVERTERS
Since this voltage acts as a bucking or boosting voltage in series with
the source, the output voltage can be effectively regulated over a wide
range of supply voltage deviation and load conditions. L4 serves to
smooth the. active feedback current to limit the pulse current duty
required of capacitor C 2 • L5 is added to provide dil dt limiting for
SCR's 1 and 2 and is not part of the active stabilization network.
J
A
B
0 1*
*0 2
L2
o...-..r-
SCRI
TI
03
"'---...
-=-OC SOURCE
SCR3
f
;;; r:CI
771\
L4
L5
"
0
A
D
L
P--
SCR4
04
L3
=-
L
r
SCR2
TRIGGER
CIRCUIT
FIGURE 13.8
VOLTAGE REf
AND FEEDBACK
CONTROL SYS.
CURRENT FEEDBACK
COMPLETE CLASS B REGULATED SINE WAVE INVERTER
13.2.2.2 Circuit Performance
Output voltage wave shapes together with main SCR current and
voltage wave shapes are shown in Figure 13.9.
Typical regulation achievable, no load to full load for a ±25%
input source variation, is ±2.5% output voltage change. The above is
based upon a 50 Hertz, 220 volt output operating from a nominal 28
volt input source. The circuit performance improves with increasing
nominal supply voltages. The circuit has been shown to have better
than the above performance levels when operated from a 220 volt input
source including variations due to a full range of leading and lagging
load power factors in addition to the no-load condition.
A final feature worthy of mention is the soft commutating duty
imposed upon the main SCR as shown by Figures 13.9(b) and (c). Note
the reverse voltage and low reapplied dv/dt during the SCR turn-off
interval.
359
SCR MANUAL
'Cal Output VDltage
(bl Main SCR Current
(el Main SCR Voltage
FIGURE 13.9 CIRCUIT OPERATING WAVEFORMS
360
CHOPPERS. INVERTERS AND CYCLOCONVERTERS.
13.2.3. Class CInverters
Typical of the Class C inverter is the well-known "McMurrayBedford Inverter." This inverter circuit is shown in Figure 13.10. It
operates as follows. Assume SCR l conducting and SCR2 blocking. Current from the DC supply Hows through the left side of the transformer
primary. Autotransformer action produces a voltage of 2Eb at the anode
of SCR2 charging capacitor C to 2Eb volts. When SCR2 is triggered,
point (A) rises to approximately 2Eb volts, reverse biases SCRb and
turns it off. Capacitor C maintains the reverse bias for the required
tum-off time. When SCR l is again triggered, the inverter returns to
the first state. It follows that the DC supply current Hows alternately
through each.side of the transformer primary producing a square-wave
AC voltage at the secondary.
(AI
FIGURE 13.10
McMURRAY·BEDFORD INVERTER
Rectifiers CRl and CR2 feed back, to the DC supply, reactive
power associated with capacitive and inductive loads. With inductive
loads, energy stored in the load at the end of a half cycle of AC voltage
is returned to the supply at the beginning of the next half cycle. Conversely with capacitive loads, energy stored in the load at the beginning
of a half cycle is returned to the supply later in that half cycle.
The feedback rectifiers are connected between the negative supply
terminal and taps on the transformer primary. In applications where
losses would not be excessive, the diodes may be returned to the SCR
anodes through small resistances as indicated by the dashed lines in
Figure 13.10. In the tap connection some energy trapped in L is fed
to the load; Use of the tap connection results in some variation of output with load power factor, but far less than with no feedback rectifiers.
361
SC~
MANUAL
Optimum values for the commutating elem,ents of a
Bedford Circuit" are:
.
'~McMurray-
to Icom
C= -==,1.7 Eb
L
where
t~ Eb
= --:--:-::-:-:-::-0.425 Icom
=
tc minimum turn-off time presented to the SCR
Icom= maximum value of load current at commutation
Inverter waveshapes for different load power factors are shown in
Figure 13.11.
The above relations define C andL such that commutation is
insured for maximum lagging load current.
lCR 0 t----'-'--L.-->----'_Ll-
LEAOING POWER FACTOR
f\
{\
LAGGING POWER FACTOR
UNITY POWER FACTOR
(NOTE, COMMUTATION INTERVALS GREATLY EXPANDED)
FIGURE 13.11
·INVERTERWAVESHAPES FOR VARIOUS LOAD POWER FACTORS
13.2.3.1 Ott Filters For Class CInverters
The Ott filter shown in Figure 13.12(a) is an extremely useful
circuit when. used in conjunction with Class C inverters. It performs
three important functions. It provides a sine wave output thus essentially eliminating the harmonic content to the load. It provides good
load regulation while at the same time maintaining a capacitive load
to the inverter over a large load range of load power factor. This
capacitive load reHected to the inverter aids SCR commutation as well
as inverter output regulation. The Smith chart shown in Figure 13.12(b)
provides the designer with a plot of filter input impedance as a function of filter load impedance, normalized to the filter design impedance, ZD'
362
CHOPPERS, INVERTERS AND CYClOCONVERTERS
o
.i.
)1
C __
II - 6ZDWD
I '" i
I'" "',w,
L2
-
=..!R..
WD
I
LOAD Z L
t!.J:.
1
o
o
ZD --FILTER DESIGN IMPEDANCE
W - - FILTER DESIGN FREQUENCY
-=Zl~
---=Z IN~
JaiN
(a)
AlEE (IEEEI CP-62-222
(II)
FIGURE 13.12
INPUT IMPEDANCE CHART
363
SCR MANUAL
While an example filter design is worked out in Section 13.2.3.2
a few examples of the use of the chart are given below to aid. in its
interpretation. The solid lines shown are those of the normalized fllter
load impedance. The radial lines being load phase angle and the circles
with center at 0, jO are load impedance magnitude values. The nor..
malized input impedance is read from the dotted lines where now· the
circles. centering on (-, jl) are capacitive phase angles and the dotted
divergent radii are normalized input impedance magnitude values. The.
values of the dotted lines are identified by being enclosed in circles to
separate them from the load set of loci.
From the example below the filter design impedance was chosen
as 15 ohms.
The following table of impedances obtained using.Figure 13.12(b)
should allow one to become proficient in its. use.
Given
load lL taL
Ohms
TD lD
30 LJI!.
20 + j20
= 28.3
Read From Smith Chart
Normalized
LJl.:.
3.1 l.;45°
46.5 1.:045°
5.5£;16°
83 £;16°
L::~5°
22.5 "-45°
1.5
45 a5°
illo
2.3 mo
34.5
t:m:
Input l
Ohms
1.88 L!5°
2
~5°
Normalized
Input lIN t alN
3
TABLE 13.2
2.15 £;-65°
32.5 £;65°
6.1 /,;;47°
91.5 l,;;;47°
12
180 L::,300
~Oo
USE OF SMITH CHART OF FIGURE 13.2(b)
By an examination of the table and the chart several advantages
of the Ott filter become apparent. First the input impedance remains
capacitive in spite of far-ranging changes in the load power factor and
impedance magnitude. Furthermore, one can easily see that as long
as the normalized load impedance magnitude exceeds 2 the input
impedance is always capacitive. The Ott filter has the further advantage of having a normalized impedance of 4.5 for open, i.e., infinite
loao impedance, unlike some filters which decrease input impedance
with increasing output impedance. Lastly the input impedance of the
filter reflects the output impedance when the output· Qecomes short
circuited, i.e., the load and the input zeroes are one and the same, thus
short circuit input current is theoretically infinite; this being ideal for
tripping protective devices under faulted load conditions. The designer
should take note that the chart of Figure 13.12(b) is valid for the filter
design of the circuit in Figure 13.12(a). 'Use of other design formula
of the same class will result in a different impedance transformation
chart. The reader is referred to Reference 3.9 of Section 13.6 for chart
construction details.
13.2.3.2 DeSign Procedure
The following is the design procedure for a Class C square wave
inverter used in connection with the Ott fllter to produce sinusoidal
voltage.
364
CHOPPERS. INVERTERS AND CYCLOCONVERTERS
Required specijicatioT18
Output voltage (Eo) - Volts (RMS)
Output power (Po) - watts
Output frequency (f) - Hz
Rated load power factor (pf)
Available DC supply (Eb) - volts
FILTER DESIGN
Load Resistance
Eo2 X pf2
RL =
P
(ohms)
o
Load Reactance
RL
XL =
y 1 - pf2 (ohms)
PI
Load Impedance
IZL! = y;";:R=-L"""2-:-+-:X;';""r,""""2 (ohms)
L ZL = ooS-1 pf (degrees)
Filter Design Impedance
IZLI
ZD ~ - 2 - (ohms)
Design Radian Frequency
roD = 27r f (radians/sec)
Filter Element Values
1
C 1 = --,-.,...,...-C2
6ZD
roD
=
1
3 Z(farads)
D roD
9Zn
(henrys)
Ll = 2 ron
Filter Input Impedance
ZIN. RIN and XIN are determined from Figure 13.12(b)
Input Voltage to Filter
E(sQ) =
~2
I
7r ZINI
(~:) (volts)
365
SCR MANUAL
INVERTER DESIGN
Transformer Turns Ratio
E(sQ)
n=-Eb
Input Power, assuming 85% efficiency
100
PI = Po X M (watts)
Average Current in SCR
Po IZINI
IAv(sCR) ~ 2 E R
b
IN
Peak Forward Voltage Across SCR's
VPK(SCR) < 2.5 Eb
From the expressions for IAv(sCR) and
choice of SCR may be made.
Peak Current in SCR's
IpK(sCR)
= 4 Eb
VPK(SCR)
a preliminary
~~
Tum-Off Time
2r
_
tC=3 yLC
Rate of Reapplication of Forward Blocking Voltage
0.85 Eb
dv/dt =
yLC
Turn-on dildt
di/dt
2Eb
=-L-
t= 0
From the preceding four relationships, Land C may be determined as follows
6 Eb to
L=-~-
'If IpK(sCR)
Choose the desired tc and
with the following
dv/dt =
3.44
IpK(sCR)
and determine L. Check dv/dt
Eb2
-=-~-
L
IpK(sCR)
If dv/dt is too high, increase L accordingly and recalculate
Now:
3 tc IpK(sCR)
C=
SwEb
IpK(sCR)'
The minimum value of L should be such as to keep the tum-on
dil dt well below specification.
366
CHOPPERS. INVERTERS AND CYCLOCONVERTERS
13.2.3.3 A400 Hz Inverter With Sine Wave"Output
The design procedure for a 400 Hz inverter with sine wave output
is given to illustrate the application of a Class C inverter used in con-
junction with the Ott filter.
Required Specifications
Output power = 360 watts
Output voltage = 120 volts (RMS)
Output frequency = 400 Hz
Rated load power factor = 0.7 lagging
Available DC supply = 28 VDC
FILTER DESIGN
Load Resistance
(120)2 X (.7)2
RL
360
=
= 20 ohms
Load Reactance
XL
T20 y l - (.7)2
=
= 20 ohms
Load Impedance
IZL! = y;;;(2:V;0:.-n)2~+~(2:n;0):n2 _ 28.3 ohms
L ZL
= cos- 1 (.7) = ..!!.-.
= 45°
4
Filter Design Impedance
Z <
D=
28.3
2
=
Choose ZD
15 ohms
Design Radian Frequency
WD = (2) (3.14) (400) = 2500 radians/sec
Filter Element Values
= 4.5 X
C2
=
1
(6) (15) (2500)
1
-:-:..,--:-~-:-:-=-:-':(3) (15) (2500)
Ll
=
(9) (15) _
-3
2 (2500) - 27 X 10 henrys
=9 X
10- 6 farads
10- 6 farads
15 -- 6 X 10-3henrys
L 2 = 2500
FiUerInputImpedance
From Figure 13.12(b) (point marked X)
ZIN = (15) 5.5 L - 16°
= 8Q-j23
RIN = 80 ohms
X1N = 23 ohms
IZINI = 83 ohms
367
SCR.MANUAL
Input Voltage to Filter
ESQ
= -y2
4 - (3.14) (83) (360)%
. 80 = 195 volts
INVERTER DESIGN
Transformer turns ratio
195
n= 28"=7
Input Power (assuming 85% efficiency)
100
PI = 360 X 85 = 424 watts
Average Cu"ent in SCR's
(360) (83)
IAV(scR) e
(2) (28) (80)
== 6.8 amps
Peak Forward Voltage Across SCR's
VPK(SCR) = (2.5) (28) = 70 volts
From the above a G-E type C141A is chosen.
Commutating Elements
The C141A has a maximum turn-off time of lOp. sec and maximum
dv/dt of 200 voltsl,usee. Choose 1:" 12 p'sec and IpK(sCR)
14 amps.
6 (28) (12)
Ii
L = (14) (3.14) = 45 X 10- henrys
=
=
Checking dvI dt,
(3.44) (790)
dv/dt = (45 X 10- 6 ) (14) = 4.3 voltsl,usee
C = (3) (12 X 10- 6 ) (14) = .75 X 10-6 farads
(8) (3.14) (28)
Tum-On dil dt
2 X 28
dildt
= 45 = 1.25 AI,usec
I
t= 0
368
CHOPPERS. INVERTERS AND CYClOCONVERTERS
..
20A
81nh
~
6mh
~
9pf
4.5,uf
1
TURNS RATIO1:1:7
.15,u1
+
GECI41A
~28V
GE CI4IA
G E IN3569
GE
In
45ph
FIGURE 13.13 A 400 Hz INVERTER WITH THE
IN~569
In
on
FILTER
NOTES:
1. Feedback rectifiers are chosen to have current and voltage
capability similar to the SCR's.
2. One ohm series resistors are used to limit power dissipation in
the feedback rectifiers.
3. The composite inverter-filter circuit for the 400 Hz inverter is
shown in Figure 13.13.
4. A suitable trigger circuit for the Class C inverter is shown in
Figure 4.50.
13.2.4 Designing a Battery Vehicle Motor·Controller Using
The Jones SCR Chopper (Class D)
13.2.4.1 Introduction
Three methods are available for controlling the voltage to, and
hence the speed of, a battery-driven DC series motor of any appreciable power:
1. A rheostat may be inserted in series with the motor. This
method has a smooth action but power is wasted in the rheostat.
2. The battery or the field winding may be switched in series or
parallel. This method is virtually lossless but the action is jerky.
3. The third method involves the use of a rapid-acting switch,
called a chopper, in series with the motor.
369
SCR MANUAL
Choppers have several modes of possible control. Figure 13.14
shows the action of the chopper for the pulse width modulation and a
combination of pulse width and frequency modulation.
FIGURE 13.14 CHOPPER WAVEFORMS
All three techniques control motor speed by varying the ratio of
switch "on" time to "off' time. At low speeds the "on" time is much
less than the "off" time. The result is that the average voltage across
the motor is low. As the "off" time is decreased so the average voltage
increases. The change in the average voltage is as smooth as the "ofF'
time may be readily adjusted by a potentiometer controlled timer. This
method combines the advantages of the two previous methods in that
both smooth control and high efficiency are achieved simultaneously.
The SCR makes an ideal switch for this chopper application.
Figure 13.15 shows a diagram complete except for the method of
turning the SCR on and off. This will be discussed later.
S2, Sa, S4 and S5 are field-reversing relays. With S2 and S5 closed
the direction is forward, whereas with Sa andS 4 closed the direction is
reverse.
This SCR chopper has a practical duty cycle ranging from about
20% to about 80%.
For the standstill state all four switches, S2, Sa, S4 and S5, are
open. When S2 and S5 are closed, and with the chopper operating at
low speed, ahout 20% of the supply voltage is applied to the motor.
This voltage may be increased to 80% of the batteQ' voltage as more
torque is required. When 80% is reached relay SI is closed applying
full voltage to· the motor and maximum torque is obtained.
370
CHOPPERS, INVERTERS AND CYCLOCONVERTERS
S,
$CRI
S3
2
.s.Eb
01
~
S4r
QARMATURE
FIGURE 13.15 BASIC VEHICLE CONNECTIONS
The diode Dl is the well known free-wheeling diode. Its purpose
is to carry the inductive current when the SCR is turned off, thus preventing high voltages appearing across the motor.
The controller to be described uses a variable-frequency constantpulse-width system.
It is capable of pulse width modulation control by changing the
trigger circuitry.
13.2.4.2 Operation of the Jones Commutation Circuit
Figure 13.16 shows the basic circuit.
FIGURE 13.16 THE JONES CHOPPER
Figure 13.16 is redrawn in Figure 13.17 to show six working
circuits representative of the basic phases of operations of the Jones
Chopper.
371
SCR MANUAL
SCRI
IO
I_--!---to TO tl
(a)
I
I
U·
01
t4 TO t6
(e)
t6 TO to
(0
FIGURE 13.17 JONES CHOPPER WORKING CIRCUITS
372
CHOPPERS, INVERTERS AND CYCLOCONVERTERS
Switches are shown instead of SCR's to indicate the conducting
state. As the switches operate during the cycle, the chopper changes
from one working circuit to another. These phases of operation represent consecutive time intervals totaling one full cycle of operation. In
the circuit waveforms of Figure 13.18 the times to, tb l::! •••• correspond with the working circuits of Figure 13.17.
The operating cycle is initiated by triggering SCRI at time to. The
bottom plate of capacitor C then starts to charge positively. Note that
in Figure 13.17(a) the amount of initial voltage (VC(to» on C is not
shown in general it will be a different value for starting than it will be
for running. The peak voltage to which C charges at tlo (Ve(tl »' depends
on (VC(to».
At time tb the bottom plate of C has resonantly charged, or "rung"
via L 2, to its peak positive voltage. Note that the peak positive voltage
is always less than, but approximately equal to, the peak negative voltage at to' In effect turning on SCRI serves to reverse the voltage on
the commutating capacitor. The peak positive voltage on C is held by
the charging diode D 2. Energy is now available to commutate SCRI.
Meanwhile SCRI is delivering power to the load through Ll as shown
in Figure 13.17(b). Note that the use of the autotransformer insures
that whenever current is delivered from the DC source to the load, a
voltage is induced in L2 in the correct polarity for charging the commutating capacitor. Thus the autotransformer measurably enhances the
reliability of the circuit.
At time t 2, SCR2 is triggered and the capacitor voltage is placed
directly across SCR I . After cessation of its reverse recovery current,
SCR, reverts to the blocking state and behaves as an open circuit.
The load current transfers to SCR2, LI and the load. This discharge
current increases until, at t a, the voltage across the capacitor is zero
and the bottom plate of the capacitor begins to swing negative (Figure
13.17(d»; forward blocking voltage is applied to SCRb and the current
in SCR2 begins to decrease. To insure continuity of inductive load current, D, begins to conduct at ta. The time duration t2 to ta is the circuit
tum-off time presented to SCRl •
373
SCR MANUAL
I
I I I
,
:-11-----·
I I
o~--~~--~~~~-----------
I
l
r-
I'
~
(BOTTOM PLATE)
I
or-----~----~~~r4I--------~~----
I
,
-' _____ J
I
I
o~----~--~~~~I~------~----
I
I
I
Notes:
SOLID LINES DEPICT OPERATION WITH DIODE 02 AS
SHOWN IN FIGURE 13.17. DOTTED LINES DEPICT OPERATION
OF CIRCUIT SHOWN IN FIGURE 13.17 WITH DIODE 02
REPLACED BY SCR3'
FIGURE 13.18
CURRENT AND VOLTAGE WAVEFORMS FOR THE JONES CHOPPER
The bottom plate of C continues to swing negative until it reaches
a peak value at time t4 when the current in SCR2 attempts to reverse
thus commutating SCR2 •
The peak negative voltage reached by C is a function of load current and inductance L 1 • It is independent of turns ratio n. The operation taking place during the period tS-t4 can be best visualized by
examining the elementary circuit of Figure 13.19.
374
CHOPPERS. INVERTERS AND CYClOCONVERTERS
FIGURE 13.19
CAPACITOR VOLTAGE BOOSTING
Prior to throwing switch S1> IL current is flowing in the inductor L.
The energy stored in L is lh LIL2. After the switch is changed from
position 1 to 2 the energy which was in L must be transferred to the
capacitor, C. Thus:
lh LIL2
lh CV0 2
L/C V 02/IL2
&
Vo= IL VLlC
(13.1)
The current IL represents the load current flowing in Ll shown in
Figure 13.17(c) prior to ta.
Since VO (t4)' in general, is greater than E b, D2 is again forward
biased and current now flows as shown in Figure 13.17(e). The capacitor voltage is now resonantly discharging down to a value less than Eb
and the blocking voltages on SCR1 and SCR2 change accordingly. Zero
voltage across L2 and SCR2 occurs when the resonating current reaches
a peak at t5' It can be seen that circuit turn-off time for SCR2 is the
time interval t4 to t5' The resonant discharging of C continues and only
ceases at time t6 when current ceases to flow in L 2 •
Improved operation can be obtained by replacing D2 by an SCR.
The operation with SCRa added is shown by the dotted lines in Figure
13.18. It accomplishes two important functions; namely, it maintains
the voltage on C to the V0(4) value, thus providing far greater stored
energy for turn-off time purposes. Secondly it allows start-up with
VO(to) charged to t-E b. This is accomplished by switching on SCR2
prior to SCR 1 to "cock" the commutating circuit. This does away with
the need for depending solely on autotransformer action to charge up
C during the first pulse.
=
=
13.2.4.3 Design Trade·Offs
The following information is required:
The battery voltage E b;
The rotor current required to provide load breakaway torque,
I oL, if current limiting is used, otherwise the locked rotor
current of the motor, 1m;
Motor time constant, t m •
Both maximum motor current and battery voltage are key variables
facing the designer who has full control of all design variables, assum375
SCR MANUAL
ing a fixed motor HP requirement. Other important and interrelated
variables are commutating capacitor size and SCR current and voltage
requirements plus autotransformer specifications. All of the above mentioned parameters and circuit component requirements are interrelated.
The rule of thumb equation generany used for circuit turn-off time
in a chopper is:
te "'"
C Eb
(13.2a)
ICL
However this does not take into account the circuit action of the Jones
Chopper which has a boost circuit which charges the commutating
capacitor voltage, V c, to values greater than Eb.ThUS Equation 13.2a
becomes:
t "'" CVc
(13.2b)
e
ICL
where Vc > Eb
At current limit the boost voltage is given by Equation 13.1. Substitution of Equation 13.1 in Equation 13.2b yields:
te = y'Ll C
(13.3)
This is valid for steady state operating conditions but not during initial
start-up. Dividing Equation 13.1 by Eb yields:
Defining
and
Vc = ICL y'-;:-L-l/~C:;---Eb
Eb
RCL = Eb/IcL
Q Vc/Eb
then
Q = Rcr. y'Ll/C
(13.4)
(13.5)
=
1
--
(13.6)
For Q»l peak voltages seen by SCR's 1, 2 and 3 are approximately
given by Q E b. This relationship is given in Figure 13.20. Thus we see
that the major design trade-offs revolve around selection of RCL, Q, Ll
and C. Where the Ll and C determine both circuit tum-off time for
SCR1 and circuit voltages.
Figure 13;21 gives a graphical representation of the major tradeoffs. From Figures 13.20 and 13.21 the major parameters for steady
state operation can be chosen. The Equations plotted in Figure 13.21
are Equations 13.3 and 13.6.
10
8
6
..,.
4
'----
~
2
0,5
1.0
2.0
4.0
6.0
-0-
FtCURE 13.20 CIRCUIT IUIIII lIlDE-oFFS
376
CHOPPERS, INVERTERS AND CYCLOCONVERTERS
1000
800
FOR Q IN RANGE OF 0.5 TO 8
600
400
Q'R/3
200
VI
>- 100
It:
I--'C= ,40
Y
,20 "
...
Z
80 15 '
l:
0
60
It:
0
i
40
I
...J
20
10
8
,-./
'.1/
',/ ,
~',
~
'/
/-, V'"
V2
/
1.5
I
V
~ I
~t>1',
/ '
J
7
V V,, l)<"
10
" .75
,
~
,V
/.3
l; ,
II ~'Y
V· 2
/
'/
,1/ '
,>(,
'k ' /
, ",
40 60 80 100
20
C - MICROFARADS
FIGURE 13.21
".5
200
I-I.15
r)
~
'oJ
400 600 1000
CIRCUIT PARAMETER RELATIONSHIPS
13.2.4.4 Design Notes
=
Assuming Ll
L 2.
Selection of SCRl
The current rating of the main SCR is determined by the motor
locked rotor current or the current limiting value, I cL. This rule of
thumb holds in practice with an adequate heat exchanger for SCRl ·
Capacitor C and L l , L2
From Figure 13.20 select a Q value based upon desired maximum
SCR and capacitor voltage ratings as well as supply voltage, E b • Determine RCL from the relationship in Equation 13.4.
From Figure 13.21 select Land C values for a tc equal to twice
the tq of the main SCR, from the intersection of tc and Q RCL product.
These values guarantee turn-off of the main SCR under steady state
conditions with a one hundred per cent safety factor. To check start-up
conditions refer to Figure 13.22.
377
SCR MANUAL
LOCKED ROTOR
FULL BATTERY VOLTAGE
....
...z
II:
II:
:::>
o
II:
~
o
~
TLR
= ~=
(MOTOR TIME CONSTANT!
TIME
FIGURE 13.22
CHOPPER START-UP
The current lA, at time T A, is given by the locked rotor conditions.
IA =(::R) (ILR)
(13.7)
For 80% voltage control TA must be 80% of the total minimum cycle
time.
(13.8)
TA "'" 4w VL l C
Commutation at IA can be checked from Equation 13.2a.
CE b
t --c IA
By substituting Equations 13.7 and 13.8 in Equations 13.2a
tc
-= ~R
4w VC/L
l
1st Commutation
or
1st Commutation
(13.9)
This equation must be satisfied for tc ~ tq main SCR.
During the second and subsequent commutations the current lb is
always less than twice the magnitude of the previous commutated current level. Since Ll and C were chosen for twice the main SCR required
turn-off time, the boosted voltage due to the previous current peak IA
will guarantee commutation at T B.
Transformer T 1
Choose a 1: 1 turns ratio.
If the transformer core has an air gap made up of the ends of the
laminations butting together with no spacer, the number of turns may
be found fr.om·the following approximate relation.
Nl = N2 =
378
~
"V"6A
CHOPPERS, INVERTERS AND CYCLOCONVERTERS
The core is assumed to be of the stacked type where A is the core cross
sectional area in square inches and LI is in p.H.
The choice of number of turns and core size must be checked
regarding the maximum flux density in the core.
=
Flux density
15Eb yLI C
NI A
It is permissible to use a core which saturates provided the voltsecond capability is chosen such that saturation does not occur before
timetl • The core is reset when reverse voltage is applied at time t 2 •
Free-Wheeling Diode DI
A rule-of-thumb for the maximum average current in DI is a
quarter of the maximum motor current, ICL or I LR . (Assume 180 0 conduction angle.)
A fast recovery diode will reduce di/dt stress on SCRI as well as
greatly reduce voltage transients that are generated by diode recovery.
Average Current in SCR 2 , SCRa, C and L2
The same average current flows in all four components.
Iavg = f (C Eb + 2 Imax yLI C) X 10- 6 amps average
Imax = ILR or ICL
where
RMS Current in LI
A rule-of-thumb for the RMS current in winding L I : half the
motor current ImoX'
Voltage Rating of SCRlo SCR 2 , Dlo and C
The peak forward and reverse blocking voltage across SCR I and
SCR2 , the peak reverse voltage across Dlo and the peak voltage across
C are found from Figure 13.20.
Voltage Rating of SCRa
N2
CPK(SCRa)
Nt VPK(SCRt)
=
SCR Dynamic Characteristic8
SCR I :
Imax
dvI dt C volts per p'S
=
E
Initial dil dt
= L: amps per p.S
Circuit Turn-off time tc obtained from Figure 13.21.
SCR2 :
VPK(SCR)1
dvI dt =
volts per p.S
yL2 C
V PK(SCRh
.. I d'/d
I mtta
1
t
Stray Inductance III Loop Formed by SCRlo SCR2 and C
Circuit turn-off time for SCR2 depends on trigger circuit timing. It may
be made several times SCRt available circuit turn-off time.
=
.
379
seR
MANUAl
SCRs:
VPK (SCR1)
Ll
R.nubber
dvI dt =
di/dt = di/dt of SCR I
Circuit turn-off time for SCRa ~ 4x tc(SCRI)
13.2.4.5 Worked Example"
Given:
= 24 volts
Eb
ILR
= 160 amps
L.n = 80 phenrys
Selection of SCRI
The GE C364 has an RMS rating of 180 amps. The SCR turn-off
time t q , is 10 p.S.
Capacitor C and LI
Assume a Q of 5. Then maximum SCR voltage ratings equal
6.3 X 24 volts or 152 volts, use 200 volt devices.
24V
RLR = 160 A = 0.15 ohms and QRLR = 0.75
From Figure 13.21 C and Ll are-respectively 40pfd and 20 ph,
allowing for a two to one turn-off time margin to enable commutation
of maximum motor current changes between subsequent cycles.
Checking commutation at first pulse, from Equation 13.9
_ 80 +20 >
tc - 4n- (.75) = 10.6 P.s
> tq
Had the basic circuit been used with a diode in place of SCRs,
the value of C would be:
(fo;:m)
C = 1.2
10 (160»)
C = 1.2 (
24
= 80 pfd
It is seen that capacitor requirements are greatly reduced by the
addition of one SCR and operation at high operating voltages.
Transformer T I:
LI
L2
20 ph
= =
Nl2A =
2~
= 3.2
If a core of 0.50 square inch cross sectional area is used, then:
.
NI"= N2 =
fIT
.50
"'V
= 3 turns
Flux Density = 15 X 24 V 40 X 20
3 X .50
15 X 24 X 89
.
15
= ~1,400 lines per square
.
mch
380
CHOPPERS. INVERTERS AND CYCLOCONVERTERS
With this flux density any of the silicon steel materials will do for the
core.
Free-Wheeling Diode DI
I L
= 4160 = 40 average amperes
+
Average Current in SC~, SCRa, C and L2
Iavg
2800 [40 X 24 + 160 X 2 y'40 X 40] X 10-6
where f = liTA
= 28.2 amperes average
RMS Current in LI
=
ILR
""2 = 80 amps RMS
Voltage Rating of SCRlo Db and C
From Figure 13.20, for Q = 5
VRRM/E B = 6.5
:. V RRM
156 volts
Voltage Rating of SCRa
V pk(SCRS)
(1) (156 volts)
= 156 volts
SCR Dynamic Characteristics
SCR1 :
160
dv I dt = 40 = 4 voltslp.Second
=
=
Initial di/dt = 156/20 = 7.8 ampereslp.second
Circuit turn-off time, t e, from Figure 13.
te =< 20 p.seconds
SCR2 :
dv I dt
= 156 voltslp.Second
y40 X 20
= 5.5 voltslp.Second
.. I di/d
Imtia
t
156
= Estlmat
'
ed at 2 ,....enry
..
1...
= 78 ampereslp.Second
Minimum te = w/2 y'L2 C p.Second
45 p.Seconds
SCRs:
dv/dt = (156/20) • 20
= 156 voltslp.Second
di/dt = di/dt SCR1
7.8 amperes!p.SeCOnd
Circuit turn"off time tc minimum of 4 X te(SCR})
=
=
=
= 80 p.Seconds
381
w
SI
FOR CIRCUIT SIMPLICITY PILOT SCR'S 5 a 6
NOT SHOWN NEXT TO SCR2 a SCRa·
RC SNUBBER NETWORKS ALSO NOT SHOWN.
R2
00
N
ON-OFF
SCR4
RI9
SCRI
CI
T2SB
SCR2
T2SB
RS
Tas
Ra
R7
R9
REVERSING
SWITCHES
RI4
D2
DUTY CYCLE
TRIGGER CONTROL
PULSE WIDTH
CONTROL
FIGURE 13.23
COMPLETE CHOPPER CIRCUIT
Parts List:
en
(")
SCR 1 - C364C
::0
SCR 2 , SCR3 - C147
Dl -lN3912
~
c::
D 2 , D3 - Z4XL18
F::
PUT 1, PUT z - D 13Tl
SCR 4 , SCR 5 , SCR6 - C106C
R1 -1K 1W
Rz-100 1W
R3, R 5 , R6 - 47 ohms, 1W
R4 -lK, 1/2W
S2 R 7 , R8 - 220 ohms, 2W
R 9 , R14 - 27K, 1W
RIO - 150K Pot (Speed Control)
Rl l , RIG - 2.2K 1W
R 1Z , R17 - \3.9K 1W
R15 - 20K Pot (Pulse Width Control)
R 13 , R 18 - 100 ohm, 1W
DI
R19 - 27 ohm, 1W
C 2 , C 3 - 0.01 /Lfd, 100 Volt
C 1 - 40 /Lfd, GE 28F5117
T1 - See Text
T 2, T 3 - Pulse Engineering Type PE 2229
S1 - On-Off & Start Control
52 - Bypass Switch
F 1 - 250 Amp, 200 Volt DC Fuse
CHOPPERS, INVERTERS AND CYCLOCONVERTERS
13.2.5 Pulse Width Modulated (PWM) Inverter
Pulse width modulation is a technique used to eliminate or reduce
unwanted harmonic frequencies when inverting DC voltage to sinewave AC. One version of the PWM technique is illustrated in Figure
13.24 where the dashed wave represents the fundamental component
which is obtained by inverting the DC supply voltage at a high frequency while modulating the width of each pulse. The result is an
output voltage that is easily filtered to produce a sinewave. The high
chopping frequency results in small lightweight magnetic components
since iron size is inversely proportional to frequency.
..-
.... "
r-
"'-
FIGURE 13.24
- - '"
PWM VOLTAGE
The PWM inverter finds use in the following applications:
• UPS, uninterruptible power supplies for stand-by power source
for computers, medical equipment, etc.
• AC motor speed control where the valiable voltage/variable
frequency capability is utilized.
• Lightweight sinewave inverter such that the high chopping
frequency results in small lightweight magnetic components.
The SCR must conduct a current waveshape comprised of pulses
which are modulating in both height and width. A computer program
is used to determine SCR current capability for any PWM waveshape
and circuit operating conditions.
13.2.5.1 The Auxiliary Commutated Inverter (Class D)
The auxiliary commutated inverter, discussed by McMurrayl.l is
one of the circuit techniques used to generate a PWM voltage. A discussion of the advantages of the auxiliary commutated PWM inverter
can be found in Reference 1.2 of Section 13.6. These include:
• High operating frequency capability resulting in small, lightweight filter components.
• Variable frequency capability.
• Excellent voltage regulation capability.
• Low no-load losses.
• Low commutation energy loss.
The basic inverter circuit of Figure 13.25 can be used in either
the half bridge or full bridge configurations of Figure 13.1. The half
bridge single phase configuration requires a center-tapped DC supply
as traded-off against the full bridge with twice as many power semiconductors. Peak to peak output voltage of the full blidge configuration
is twice that of the half bridge, eliminating the need for voltage trans383
SCR MANUAL
formation in some applications. The basic inverter circuit can be' used
as a building block for the three-phase circuit of Figure 13.1.
+
r----- -----,
II
I
I
I
I
I
I
I
SCR,
D,
I
I
I
V'''''''''''-II~~-+---+I-oOUTPUT
'--_. . . ._-=
I
D21
ILI _____ _ _____ JI
FIGURE 13.25 BASIC CIRCUIT AUXILIARY COMMUTATED INVERTER
A detailed discussion of the theory of operation is contained in
Reference 1.1 of Section 13.6. Briefly, SCR 1 is the main SCR whose
task it is to deliver current from the DC supply to the load. SCRu is the
commutation SCR which forces the current in the main SCR to' zero
by an impulse current discharge of the L-C components. By controlling
the ratio of ON to OFF time of the main SCR, SCRlo the desiredPWM
voltage will appear at the output.
Waveshapes of current and voltage for the main and auxiliary
SCR's are shown in Figure 13.26.
f\
0
+2.3Ed
F
I
0
-1.3E d
ISCRI
vSCRI
o
------~----------~~--------~\----------L_______
+
E
d
._
J
0 _
__
_._
_
+Ed----""""""
VLOAD
0
-Ed
LI_ _ _ _--J
-
FIGURE 13.2& WAVESHAPES FOR FULL BRIDGE CIRCUIT
384
r_
vSCRIA
fl
~
ISCRIA
L
CHOPPERS, INVERTERS AND CYCLOCONVERTERS
13.2.5.2 Desiln Notes
The auxiliary SCR, SCRlA, must be capable of conducting a high
peak, narrow pulse current wave. This indicates the need for a fieldinitiated gate structure. The reader is referred to Chapter 1 for a discussionof the FI gate and its derivatives.
To achieve adequate commutation of the main SCR, the L-C discharge current must exceed the load current, I L , for an interval te
which is longer than the turn-off time of the main SCR, tqo The
optimum impulse current waveshape, that requiring the least amount
1.5 IL , as in Figure 13.27. The width
of energy, occurs for Ip
of the commutating current pulse, t pw , is determined by the resonant
components.
=
2n7r
= 2n-f = - - = - 2 tpw
tpw
tpw
7r yLC
W
=
The commutation components can then be expressed in terms of
turn-off time of the main SCR.
VLC = .6tq
Therefore an SCR with low tq will be chosen in order to minimize the
commutation components, Land C.
The ratio of peak commutation capacitor voltage, E c , and Ip can
be approximated as
~~ =~:
The commutation capacitor and inductor are then determined as
.6 tq Ip
C=--
Ec
Ec)2
.6tq Ec
L= ( C=-Ip
FIGURE 13.27
1]1
COMMUTATION CURRENT PULSE
385
SCR MANUAL
The blocking voltage requirement of the commutation thyristor,
SCRlA, is dependent on the capacitor voltage, Ec.
= Ed + [ X IL sin wt2 - (Ed - E
Ed = DC supply voltage
X = VL/VC
IL = Load current
w = Approximately (LC)-%
t2 = Portion of commutation interval
Ec
where
1)
J
cos wt2 exp [ -
;~ ]
El = Initial capacitor voltage at start of time ~
Q =X/R
The above equation for Ec is derived and plotted in Reference 1.1
of Section 13.6. As an approximation the peak capacitor voltage can be
expressed as
Ec 2.5 Ed
Actual performance can then be experimentally determined.
13.2.5.3 Design Example
Required output: no volts RMS
400 Hertz sinewave.
400 A peak load current
The main SCR is chosen for fast turn-off time in order to minimize
commutation component size. The C395 is capable of conducting the
load current with a maximum turn-off time of 20 microseconds, under
severe test conditions.
Choose chopping frequency of 2 kHertz
=
2000 Hz
400Hz
= 5.1
.
Commutation components are determined by the impulse current
required.
Ip = 1.5IL
= 600 amperes peak
For turn-off time of the main SCR of 20 microseconds
tpw 2 (20 p.Seconds)
= 40 p.Seconds
The C358 SCR is chosen to meet the current requirements of a
600 amp peak half sinewave of 40 microsecond pulse width operating at 2 kiloHertz.
=
initial dil dt
= wfpwIp
= 50 amperes per microsecond, repetitive
The DC supply voltage is estimated to be
now
Ed
=2 V 2
= 125 volts DC
386
CHOPPERS, INVERTERS AND CYCLOCONVERTERS
The commutation SCR blocking voltage is chosen to be three
times the supply voltage. Peak capacitor voltage, Eo
Eo 2.5 Ed
= 300 volts
The commutation capacitor can then be determined.
=
C
= .6 to Ip
Eo
_ .6 (20 p,S) (600 A)
300 V
24 ftFarad
From the capacitor selection chart in Chapter 5 select 28F5116,
a 30 pFarad capacitor. The commutation inductance size is
calculated
=
L
= EcIp
2
(C)
(300 V)2
= (600 A)2 (30 pF)
L = 7.5 ftH
RC snubber network values depend on stray inductance and
source impedance, thus may be determined experimentally.
Fast recovery A396 diodes are used in order to minimize reverse
recovery currents, power dissipation, and voltage transients.
SCR2
SCRI - SCRA = C3950
SCRIA - SCR4A = C3580
01 - 04
L
= A3960
C
= 30)JF 28F5116
=~5pH
SNUBBER CIRCUITS NOT SHOWN
FIGURE 13.28 SINGLE PflASE CIRCUIT
13.3 INVERTER ACCESSORIES
In practical applications of inverters it is often necessary to modify
the design to accommodate one or more of the following requirements:
1. The ability to operate into inductive loads
2. Over-current protection
3. Open-circuit operation
4. Sine wave output
5. Regulated output
387
SCR MANUAL
13.3.1 The Ability to OperatelRto Inductive Loads
When an inverter sees a reactive load as opposed to a purely
resistive load several changes occur in the operation of the inverter.
Without attention, a reactive load can cause high voltage transients to
exist in the inverter resulting in loss of efficiency and power and
jeopardizing the components.
Consider Figure 13.29. Assume that SCR 1 is conducting. Current
is flowing in the primary of the transformer as shown by arrow "a"
and in the load by "b". When SCR 1 is turned off, current "b" still
needs to flow. If no path were provided in the primary, the voltage
would rise excessively.
br2:=::::J
FIGURE 13.29 FLOW OF REACTIVE CURRENT
A convenient means of providing a current path is to place a
diode across SCR2 • Now current "c" can flow and this is magnetically
the same path as current "a". The dv/dt that the circuit applies to the
SCR is greatly increased. Figure 13.30 shows the effect of a diode
across the SCR on the voltage waveform. In Figure 13.30(b) it is seen
that the voltage across the SCR is held at a low negative value while
current is flowing through the diode. When the diode ceases to carry
current, the voltage across the SCR suddenly snaps up to a high value.
As the rise time is commonly less than 1 p$, the value of dv/dt can be
very high. Where possible, it is preferable to avoid placing the diode
directly across the SCR. The circuit of Figure 13.10 for. example shows
how an inductor can be used between the SCR and the diode. By this
means both the high values of dv I dt and the low amount of reverse
voltage can be avoided.
388
CHOPPERS. INVERTERS AND CYCLOCONVERTERS
VSCR
Or--r--r------------
Ol------.::=-r------TIME
la)WITHOUT DIODE ACROSS SCR
(Ii WITH DIODE ACROSS SCR
FIGURE 13.30 VOLTAGE WAVEFORM ACROSS THE SCR
13.3.2 Overcun-ent Protection
If the load current in an inverter is increased beyond the rated
output, some means must- be provided for the protection of the components. The following methods may be considered.
13.3.2.1 Fuses and Circuit Breakers in the DC Supply
This, the most obvious of steps, has the advantage of simplicity.
It is however necessary to match the overload capabilities of the SCR
with the current-time rating of the fuses or circuit breakers. Thus the
I 2t rating of the SCR must be greater than that of the fuse. This is
complicated by the fact that the I 2t rating of the SCR drops substantially during the SCR turn-on time, and fuses or circuit breakers do
not afford very good protection in this short time.
Another snag is the location of the fuse in the DC supply. Invariably a ripple current due to the load current flows in the DC supply.
Thus the fuse will see a relatively high RMS current and may, in the
case of high frequency inverters, have to be derated because of skin
effect. If on the other hand a large filter capacitor is placed between
the fuse and the inverter to carry the ripple current then the fuse does
not isolate the SCR from the energy in the capacitor.
13.3.2.2 Current Limiting by Pulse-Width Control
The inverter components may be protected by sensing the output
current and using this information to narrow down the pulse width
when the output current exceeds the rated value. With a very heavy
389
SCR MANUAL
load the current pulses then become narrow and have a high amplitude.
The circuit is then liable.to present short values of turn-off time and
high values of dil dt to the SCR. If the load is distributed to more than
one piece of apparatus, there may not be enough current in the. case
of a current limited supply to blow the local fuse where a short circuit
occurs.
13.3.2.3 Current Limiting by LC Resonance
The bridge circuit in the output lead of the inverter in Figure
13.31 is in series resonance at the output frequency. If the Q of the
capacitors and inductors is high, the overall efficiency of the inverter
will not be appreciably changed.
In the event of a current overload a fast acting switch is connected
between points A and B. The bridge circuit then becomes a parallel
resonant circuit at the operating frequency and the impedance to the
load current becomes very high.
The fast-acting switch may be either a saturating reactor or one of
the forms of SCR AC switches described in Chapter 8.
A
Ell
LOAD
L -________________
FIGURE 13.31
~
CURRENT LIMITING BY RESONANCE
13.3.2.4 Current Limiting in Class ACircuits by Means of
Series Capacitors
Class A circuits such as Figure 13.32 can be made current limiting
by connecting a capacitor C 1 in series with the load R. See Figure
13.32. The value of capacitor is chosen so that, when the load is
shorted,· the resonant frequency of the LC·· circuit is still appreciably
greater than the triggering frequency. Figure 13.33 shows a typical
curve of load current versus load voltage.
390
CHOPPERS. INVERTERS AND CYClOCONVERTERS
c
R
FIGURE 13.32 CURRENT LIMITING IN A CLASS A INVERTER CIRCUIT
140
- ---
120
100
~
~
""
'\
1\
\
40
20
o
3
4
5
6
8
9
LOAD CURRENT IN AMPS (RMSI
FIGURE 13.3.3 LOAD REGULATION CURVE OF A CURRENT LIMITING CLASS A INVERTER
13.3.3 Sine·Wave Output
Most applications of DC to AC inverters prefer a sine-wave rather
than a square wave output. In conHict with this we are faced with the
fact that the SCR is essentially a switch and switching a battery gives
square waves. In fact the great efficiency with which SCR inverters
391
10
SCR MANUAL
operate is mainly due to the fact that the SCR switches at high speed
from the fully-off to the fully-on mode.
Sine-wave output-waveforms may be obtained from SCR inverters
by the following approaches:
1. Resonating the load
2. Harmonic attenuation by means of an LC filter
3. An LC filter plus optimum pulse-width selection
4. Synthesis by means of output voltage switching
5. Synthesis by control of the relative phase of multiple inverters
6. Multiple pulse width control
7. Selected harmonic reduction
8. Cycloinversion
13.3.3.1 Resonating the Load
The waveform in the load may be made sinusoidal by inserting
the load in a resonant circuit of a Q high enough to achieve the desired
harmonic content. A typical circuit using this approach. is found in
Class A inverters. Owing to the large size of the LC components this
circuit only becomes attractive above about 400 Hz.
13.3.3.2 Harmonic Attenuation by Means of an LC Filter
This filter can take many forms. The most attractive is that by Ott
described in Reference 3.9 of Section 13.6. The circuit is shown in
Figure 13.12(a). The Ott filter has the following very desirable
characteristics;
1. Good voltage transfer characteristics.
2. Attenuation independent of the load.
3. The input impedance can be designed to be capacitive over the
working load range.
For details see the Class C inverter example in Sections 13.2.3.1
and 13.2.3.2.
13.3.3.3 An LC Filter Plus Optimum Pulse Width Selection
The requirements of the LC filter can be appreciably reduced by
using a narrower pulse width than 180°. Thus a 120° pulse has zero
third harmonic distortion. See Figure 13.34.
392
CHOPPERS, INVERTERS AND CYCLOCONVERTERS
Ihd
D
-~-D
l
PULSE WI DTH IN DEGREES
8
COURTESY OF LTV MILITARY ELECTRONICS DIVISION, LINGTEMCO-VOUGHT, INC., DALLAS, TEXAS.
FIGURE 13.34 HARMONIC CONTENT VERSUS PULSE WIDTH WITH RECTANGULAR WAVEFORMS
13.3.3.4 Synthesis by Means of Output-Voltage Switching
The output from an inverter is coupled through a transformer
(Figure 13.35) to the load via SCR "tap switches". The appropriate
SCR is triggered to give the output waveform shown in Figure 13.36.
The inverter operates, in this case, at five times the output frequency.
This waveform is easily filtered to give a good sine wave output.
OUTPUT
FIGURE 13.35 OUTPUT VOLTAGE' SWITCHING CIRCUIT
393
SCR MANUAL .
3
3
5
5
5
6
6
4
4
2
FIGURE 13.36
WAVEFORM WITH OUTPUT VOLTAGE SWITCHING
13.3.3.5 Synthesis by Controlling the Phase Relationship of Multiple
Inverters
The basis of this method of hannonic reduction is to add the
outputs of a multiplicity of inverters to form a quasi sine wave which
has no low-order harmonic component. The remaining high-order harmonics are easily IDtered.
Figure 13.37 shows an outline of a three phase bridge inverter
circuit. The voltages across outputs a, band c are shown in Figure
13.38(a) and the line-to-line voltage across the output transformer is
shown in Figure 13.38(b). If more phases are used the steps in the
waveform become smaller giving an even lower harmonic content.
FIGURE 13.37 THREE PHASE BRIDGE INVERTER CIRCUIT
v.
WCT
,,"WCT
"eTOCT
I
L
I
~
(a)
FIGURE 13.38 THREE PHASE BRIDGE WAVESHAPES
394
CHOPPERS. INVERTERS ANDCYClOCONVERTERS
Vo TO b
(b)
FIGURE 13.38
THREE PHASE BRIDGE WAVESHAPES
13.3.3.6 Multiple Pulse Width Control
This method of achieving a sine wave output is obvious from Figure 13.39. This waveform may be obtained from a bridge circuit
(Figure 13.37). One pair of SCR's is triggered and turned off with
various pulse widths to form the positive half cycle and then the other
pair is operated similarly for the negative half cycle.
FUNDAMENTAL COMPONENT
FIGURE 13.39
OUTPUT WAVEFORM USING MULTIPLE PULSE WIDTH CONTROL
13.3.3.7 Selected Harmonic Reduction
The circuit in Figure 13.14 is triggered so as to give a load waveform as shown in Figure 13.40. With precise control of the pulse
widths the output wave-shape can be made low in 3rd and 5th harmonic content.
The advantages of this method over other methods of synthesis
are:
1. The fundamental output may be varied from zero to maximum
amplitude without re-introducing the harmonic voltages.
2. A three-phase circuit using only twelve SCR's can eliminate
all the harmonics below the eleventh while still being able to
control the fundamental frequency from zero to maximum.
3. The triggering circuitry is considerably simplified.
395
SCR MANUAL
FIGURE 13.40
OUTPUT WAVEFORM WITH SELECTED HARMONIC REDUCTION
13.3.3.8 The Cycloinverter
A cycloinverter consists of an -inverter, operating at about ten
times the desired output frequency, to which is coupled a cycloconverter (see Section 13.5) producing the desired output frequency
waveform and amplitude.
13.3.4 Regulated Output
Most inverter customers specify that the inverter output beregulated both for input voltage and load current variations.
The inverter designer has three choices:
1. To regulate the supply voltage to the inverter
2. To regulate within the inverter
3. To regulate the output of the inverter
13.3.4.1 Supply-Voltage Regulation
If the supply is a battery, fuel cells or some other DC supply, then
the pre-regulation takes the form of a regulated DC to DC converter,
the logic for the DC to DC converter coming from the output of the
inverter, Figure 13.41.
The DC to DC regulated supply can take many forms. If the
inverter supply is from a rectified· AC line, then pre-regulation can be
achieved by substituting phase controlled SCR's for the rectifier diodes.
The logic for the triggering circuits is again supplied from the output
of the inverter, Figure 13.42. Phase controlled rectifiers are discussed
in Chapter 9.
±
T
DC TO DC
REGULATED
SUPPLY
I
FIGURE 13.41
396
L
DC TO AC
INVERTER
LOAD
I
DC SUPPLY VOLTAGE REGULATION
CHOPPERS. INVERTERS AND CYCLOCONVERTERS
AC
LINE
PHASE
INVERTER
RECTIFIER
FIGURE 13.42
LOAD
I
I
I
(]
DC TOAC
CONT1IOLLEII
AC SUPPLY VOLTAGE REGULATION
13.4 PULSE MODULATOR SWITCHES
The semiconductor switch in a pulse modulator circuit is required
to discharge a capacitor extremely fast, through a noninductive path.
A specially characterized pulser SCR is needed to conduct the resulting high amplitude, fast-rising current pulse. Pulse widths vary from
.1 ,.,.second to 10 ,.,.seconds in applications such as radar modulators and
laser pulsers, as. in the circuit of Figure 13.43.
If tum-off time is not a critical characteristic of the SCR because
of the low repetition rate, advantage can be taken of the SCR design
trade-off existing between tum-off and tum-on capabilities. See Chapter 1 for discussion of gold doping. Pulser SCR's designed specifically
for superior tum-on characteristics are capable of fast tum-on, low
tum-on power dissipation, and extremely high di/dt capability.
Airborne radar, with its higher frequency requirements, necessitates an SCR with both fast tum-on and fast tum-off capabilities. High
frequency SCR's such as the Cl4I and C144 make ideal pulse modulator switches for high frequency applications.
In order to achieve high dil dt capability the gate of the SCR
should be driven as hard as the ratings permit. The rise time of the
gate pulse should be less than .1 ,.,.second.
CHARGING
CHOKE
PULSER
SCR
~
LASER
_ - - - - - -.....- - - - - - - - - ' DIODE
FIGURE 13.43 LASER PULSER CIRCUIT
13.5 CYCLOCONVERTERS
A cycloconverter is a means of changing the frequency of alternating power using controlled rectifiers which are AC line (Class F)
commutated. The cycloconverter is thus an alternative to the frequency
changing system using a rectifier followed by an inverter.
397
SCR MANUAL
13.5.1 Basic Circuit
The method of operation is readily understood from Figure 13.44.
A single-phase full-wave rectifier circuit is equipped with two sets of
SCR's, which would give opposite output polarities. Thus if SCR 1 and
SCRI
FIGURE 13.44
SINGLE PHASE CYCLOCONVERTER
SCR 2 are triggered the DC output would be in the polarity shown on
the left side of Figure 13.45. If SCR3 and SCR 4 were triggered instead,
the output polarity would be reversed as shown. Thus, by alternately
triggering the SCR pairs at a frequency lower than the supply frequency a square wave of current would How in the load resistor. A filter would be needed to eliminate the ripple.
In order to produce a sine wave output, the triggering of the individual SCR's would have to be delayed by varying degrees so as to
produce the waveform shown in Figure
f\f\f\f\f\f\f\f\
INPUT
V \TV vVVV
OUTPUT
OUTPUT AFTER FILTERING
FIGURE 13.45
398
~---~
SINGLE .PHASE CYCLOCONVERTER WAVEFORMS
CHOPPERS, INVERTERS AND CYClOCONVERTERS
13.5.2 Polyphase Application
The SCR cycloconverter is important in two applications. In the
variable-speed, constant-frequency system an alternator is driven by
a variable-speed motor such as an aircraft engine, yet the required
output must be at a fixed and precise frequency such as 400 Hz.
A second application is where the required output must be variable
in both frequency and amplitude for driving an induction or synchronous motor. This makes possible variable speed brushless motors which
could for example be used to drive the wheels of vehicles operating in
difficult environments.
Due to the advantages in AC motor design, most of the cycloconverter systems are polyphase. Figure 13.46 shows a typical schematic
(excluding the trigger circuit).
GENERATOR
FIGURE 13.46
THREE PHASE CYCLOCONVERTER CIRCUIT
13.6 SELECTED BIBLIOGRAPHY
1.0 Inverters - General
1.1 SCR Inverter Commutated by an Auxiliary Impulse, W.
McMurray, 1964 Proceedings of the INTERMAG Conference.
1.2 Inverter Commutation Circuits,. A. J. Humphrey, IEEE/
IGA Conference, October 1966, pp. 97-108.
1.3 Principles of Inverter Circuits, (Book), B. D. Bedford and
R. G. Hoft, Wiley 1964.
1.4 An SCR Inverter With Good Regulation and Sine Wave
Output, N. Mapham, IEEE/IGA Conference Record, October 1966, pp. 451-472, or IEEE/IGA Proceedings, AprilMay 1967.
399
SCR MANUAL
2.0 Choppers - DC Motor Speed Control
.
2.1 Design Analysis of Multi-Phase DC-Chopper Motor Drive,
E. Reimers, IEEE/IGA Conference Record, October 1970,
pp. 587-595.
2.2 High Efficiency, High Power, Load Insensitive DC Chopper
for Electronic Automobile Speed Control, V. Wouk, IEEE/
IGA Conference Record, October 1969, pp. 393-402.
2.3 The Control of Battery Powered DC Motors Using SCR's
in the Jones Circuit, J. C. Hey, N. Mapham, IEEE International Convention Record, Part 4, 1964.
2.4 Analysis of Energy Recovery Transformer in DC Choppers
and Inverters, S. B. Dewan, D. L. Duff, Workshop on
Applied Magnetics Proceedings, May 1969, 11-3.
2.5 Analysis of Thyristor DC Chopper Power Converters Including Non-Linear Commutating Reactors, W. McMurray,
Workshop on Applied Magnetics Proceedings, May 1969,
11-2.
2.6 Thyristor (SCR) Chopper Control System for Transportation
Equipment, E. F. Wiser, 1968 IEEE/IGA Conference
Record, pp. 471-482.
2.1 The Use of Thyristors for the Control of a DC Traction
Motor Operating From a 600 Volt Line Supply, J. Beasley,
G. White, lEE Conference No. 11, Power Applications of
Controlable Semiconductor Devices, November 1965, pp.
187-195.
2.8 Variable Pulse Width Inverter, D. Jones, Electronic Equipment Engineering, November 1961, pp. 29-30.
2.9 A DC Motor Servo Controlled by a Pulse Width Modulated
Inverter, P. Bowler and W. K. O'Neill, Direct Current,
Vol. 2, No.1, February 1911.
2.10 Guidelines on Adaptation of Thyristorized Switch for DC
Motor Speed Control, Z. Zabar and A. Alexandrovitz, IEEE/
IGA Conference Record, February 1970, p. 10.
2.11 "Development of and Operational Experience With a High
Powered DC Chopper for 1500 Volt DC Railway Equipment," C. E. Band and J. H. Stephens, IEEE Publication
#53, Conference on Power Thyristors and Their Applications, Part 1, May 1969, pp. 271-288.
2.12 "The GM High Performance Induction Motor Drive System," P. D. Agarwal, IEEE on Power Apparatus and
Systems, Vol. PAS-88 #2, February 1969, pp. 86-93.
2.13 "Part II - The Application of the Separately Excited DC
Traction Motor to DC and Single Phase AC Rapid Transit
Systems and Electrified Railroads," R. A. Van Eck, IEEE
Conference Record, IGA 1969, pp. 229-237.
2.14 "Thyristor (SCR) Chopper Control for Transportation Equipment," R. A. Zeccola and E. F. Weiser, IEEE Transactions
on IGA, Vol. IGA-5 #4, July/August 1969, pp. 470-475.
2.15 "Design Considerations Pertaining to a Battery Powered
Regenerative System," B. Berman, IEEE/IGA Conference
Record 1911, pp. 341-346.
400
CHOPPERS. INVERTERS AND CYCLOCONVERTERS
2.16 "All Solid State Method for Implementing a Tractive Drive
Control," B. Berman, IEEE/IGA Conference Record 1971,
pp. 341-346.
2.17 "DC Chopper With High Switching Reliability and Without the Limitation of the Adjustable Mark-Space Ratio,"
Hokalis and J. Lemmrich, lEE Proceedings Conference on
Power Thyristors and Their Application - London, May 6-8,
1969, pp. 208-215.
2.18 "Battery Powered Regenerative SCR Drive," B. Berman,
IEEE Conference Record 1970, IGA Group Meeting, pp.
657-662.
2.19 "Controller Induced Losses in Electric Vehicle Drives," C. J.
Amato, IEEEIIGA Conference Record, pp. 457-469.
2.20 "DC Choppers for Railway Applications," C. Jauquet, J.
Gouthiere and H. Hologne, lEE Proceedings Conference on
Power Thyristors and Their Applications, London, May 6-8,
1969, p. 289.
3.0 UPS, Uninterruptible Power Systems
3.1 Application of Static Uninterruptible Power Systems to
Computer Loads, A. Kusko, F. E. Gilmore, IEEEIIGA Conference Record, October 1969, pp. 635-641.
3.2 Static Inverter Standby AC Power for Generating Station
Controls, J. D. Farber, et aI, IEEE Transactions on Power
Apparatus Systems, Vol. PAS-87, No.5, May 1968, pp.
1270-1274.
3.3 Wide Range Impulse Commutated, Static Inverter With a
Fixed Commutation Circuit, F. G. Turnbull, IEEE Conference Record of IGA, October 1966, pp. 475-482.
3.4 A True No-Break, Off-Line UPS, L. J. Lawson, IEEE/IGA
Conference Record, 1967.
3.5 Large Static UPS, Industrial Static Power Converter, C. W.
Flairty, A. E. Relation, IEEE Industrial Static Power Converter Conference, November 1965.
3.6 Pulse Width Modulated Inverters for UPS Applications,
W. V. Peterson, E. J. Yohman, IEEE Transactions on Industrial Electronics and Control Instrumentation, Vol. IE CI-17,
No.4, June 1970, pp. 339-345.
3.7 UPS Systems for Generating Stations, C. G. Helmick,
69CP731. IEEE Summer Power Meeting, Dallas, Texas,
June 1969.
3.8 Fundamentals of PWM Power Circuit, L. .T. Penlowski and
K. E. Pruzinsky, IEEE/IGA Conference Record 1970, pp.
669-678.
3.9 A Filter For Silicon Controlled Rectifier Commutation and
Harmonic Attenuation in High Power Inverters, R. R. Ott,
Communications and Electronics, May 1963, pp. 259-262.
3.10 A 50-kva Adjustable-Frequency 24-Phase Controlled Rectifier Inverter, C. W. Flairty, Direct Current, December 1961,
pp. 278-282.
3.11 A High Power DC-AC Inverter With Sinusoidal Output,
G. Salters, Electronic Engineering, September 1961, pp.
586-591.
401
SCR MANUAL
3.12 Successful Uninterruptible Power Systems for Computers,
R. Morrison Renfew, IEEE/IGA Conference Record, Vol.
IGA 5, November 1969, p.693.
3.12 A Filter For Silicon Controlled Rectifier Commutation and
Harmonic Attenuation in High Power Inverters, R. R. Ott,
Communications and Electronics, May 1963, pp. 259-262.
3.13 "UninterruptiblePower for Critical Loads," A. E. Relation,
IEEE Transactions on Industry and General Applications,
Vol. IGA-5 #5, September/October 1969, pp. 582-587.
3.14 "Inverter for Uninterruptible Power Supplies Will Subcycle
Fault Clearing Capability," Loran H. Walker, IEEE/IGA
Conference Record 1971, pp. 361-370.
3.15 "Designing for System Reliability in Large Uninterruptible
Power Supplies," C. G. Helmick, IEEE/IGA Conference
Record 1971, pp. 371-384.
3.16 "UPS Systems for Critical Power Supplies," A. E. Relation,
Solids tate Controls, Inc., pp. 877-884.
3.17 "A Three Phase 250 KVA No Break Power Supply With
Current Limiting Filter," J. Weaver, A. M. Eccles and W. P.
Kelham, lEE Proceedings Conference on Power Thyristors
and Their Applications, London, May 6-8, 1969, p. 339.
3.18 "High Power Thyristor Inverters for Essential Service,"
R. A. Hamilton, J. L. Fink and J. F. Shedlock, lEE Proceedings Conference on Power Thyristors and Their Applications, London, May 6-8, 1969, p. 305.
3.19 "Non-Break AC Power Source Switching Equipment," H.
Goshima, lEE Proceedings Conference on Power Thyristors
and Their Applications, London, May 6-8, 1969, p. 193.
3.20 "Uninterruptible Power Supply (UPS) Systems for Generating Stations," C. G. Helmick, 69CP731 IEEE Summer
Power Meeting - Dallas, Texas, June 22-27, 1969.
4.0 dv/dt and dildt
4.1 Design of Snubber Circuits for Thyristor Converters, J. B.
Rice, IEEE/IGA Conference Record, October 1969, pp.
485-490.
4.2 Analysis and Design of Optimized Snubber Circuits for
dv/dt Protection in Power Thyristor Applications, S. J. WU,
presented at IEEE/IGA Conference, November 1970 and
available as Publication 660.24* from General Electric Company, Syracuse, New York.
4.3 The Rating and Application .of SCR's Designed for Switching at High Frequencies, R. F. Dyer, Application Note
660.13,* General Electric Company, Syracuse, New York.
4.4 Voltage Transient and dv/dt Suppression in Thyristor
Bridges, J. Merrett, Mullard Technical Communications,
No. 92, March 1968.
4.5 Improved Performance for Solid State Inverters Via the
Amplifying Gate SCR, J. C. Hey, IEEE Cleveland Electronics Conference Record, April 1969.
402
CHOPPERS. INVERTERS AND CYClOCONVERTERS
4.6
"Optimum Snubber for Power Semiconductors" by W.
McMurray, IEEE/IGA Conference Record 1971, pp.
885-893.
5.0 Cycloconverter
5.1 The Practical Cycloconverter, L. J. Lawson, IEEE/IGA
Conference Record, October 1966, pp. 123-128.
5.2 AC to AC Frequency Converter Using Thyristors, S. B.
Dewan, P. P. Diringer, NEC, Chicago, October 1966.
5.3 High Frequency Power Conversion (PNPN High to Low
Frequency Converter), P. W. Clarke, AlEE Conference
Paper 62-335, 1962.
5.4 Static Adjustable Frequency Drives, J. W. Nims, IEEE
Transactions on Applications and Industry, May 1963, pp.
75-79.
5.5 Frequency-Changer Systems Using the Cycloconverter
Principle, R. A. Van Eck, May 1963, pp. 163-168, AIEE
Transactions Applications and Industry.
5.6 A Polyphase, All Solid State Cycloconverter, G. J. Hoolboom, IEEE Conference Paper 63-1040, October 1, 1963.
5.7 Cycloconverter Adjustable Frequency Drives, J. C. Guyeska,
H. E. Jordan, IEEE Textile Industry Conference, October
1-2,1964.
5.8 Static AC Variable Frequency Drive, J. C. Guyeska, H. E.
Jordan, IEEE Conference Paper CP 64-39l.
5.9 Static Frequency Converter, L. J. Lawson, Proceedings of
the 19th Annual Power Sources Conference, May 18-20,
1965, pp. 135-137.
5.10 Precisely Controlled 3-Phase Squirrel Cage Induction
Motor Drives for Aerospace Applications, L. J. Lawson,
IEEE Transactions on Aerospace, June 1965, pp. 93-97.
5.11 A Variable-Speed Constant Frequency Generating System
for a Supersonic Transport, K. M. Chirgwin, IEEE Transactions on Aerospace, June 1965, pp. 387-392.
5.12 Sub-Ripple Distortion Components in Practical Cycloconverters, C. J. Amato, IEEE Transactions on Aerospace, June
1965, pp. 98-106.
5.13 Analog Computer Simulation of an SCR as Applied to a
Cycloconverter, C. J. Amato Proceedings of NEC, 1965,
Vol. 21, pp. 933-937.
5.14 Precise Control of a 3-Phase Squirrel Cage Induction Motor
Using a Practical Cycloconverter, W. Slabiak, L. J. Lawson,
Proceedings of NEC, 1965, Vol. 21, pp. 938-943.
5.15 Thyristor Phase-Controlled Converters and Cycloconverters,
(Book), B. R. Pelly, Wiley 1971.
5.16 Variable Speed With Controlled Slip Induction Motor, C. J.
Amato, IEEE Industrial Static Power Conversion Conference Record, November 1, 1965, pp. 181-185.
5.17 Optimizing Control Systems for Land Vehicles, W. Slabiak,
IEEE Industrial Static Power Conversion Conference Record, November 1, 1965, pp. 186-189.
403
SCR MANUAL
5.18 AC Motor Supply With Thyristor Converters, L. Abraham,
J. Forster, G. Schliephake, IEEE Industrial Static Power
Conversion Conference Record, November 1, 1965, pp.
210-216.
5.19 The Application of a Cycloconverter to the Control of Induction Motors, P. Bowler, Conference on Power Applications
of Controllable Semiconductor Devices, London, England,
November 10-11, 1965, lEE Publication No. 17, pp.
137-145.
5.20 "A Method for Harmonic Analysis of Cycloconverters," S. B.
. Dewan and M. D. Kankam, IEEE/IGA Transactions, Vol.
IGA-6 #5, September/October 1970, pp. 455-462.
5.21 "An AC Equivalent Circuit for a Cycloconverter," C. J.
Amato, IEEE Transactions on IGA, Vol. IGA-2 #5, September/October 1966, pp. 358-362.
5.22 "A Synchronous Tap Changer Applied to Step-Up Cycloconverters," W. R. Light, Jr. and E. S. McVey, IEEE Transactions on IGA, Vol. IGA-3 #3, May/June 1967, pp.
244-249.
5.23 "A Single-Phase/Polyphase Converter," J. H. Parker, W. C.
Beattie, IEEE/IGA Conference Record 1971, pp. 43-49.
5.24 "A Modified Cycloconverter for use with High Frequency
Sources, J. E. Jenkins, lEE l'roceedings Conference on
Power Thyristors and Their Applications, London, May 6-8,
1969, pp. 313-319.
5.25 "Characteristics of the SCR Cycloconverter Power Stage,"
E. J. Yohman, 1968 WESCON Session 26.
5.26 "Cycloconverter Control Circuits," T. M. Hamblin, T. H.
Barton, IEEE Conference Record of 1970, Fifth Annual
IGA Meeting, October 1970, pp. 559-571.
5.27 "Harmonic Analysis of AC to AC Frequency Converter,"
S. B. Dewan, P. B. Biringer, G. J. Bendezak, IEEE Transactions on Industry and General Applications, Vol.· IGA-5
#1, January/February 1969, pp. 29-33.
5.28 "Cycloconverter Control of the Doubly Fed Induction
Motor," W. F. Long, N. L. Schmitz, IEEE Transactions on
IGA, Vol. IGA-7 #1, January/February 1971, pp. 95-100.
5.29 "Precise Control of a Three-Phase Squirrel-Cage Induction
Motor Using a Practical Cycloconverter," W. Slabiak and
L. J. Lawson, IEEE Transactions on IGA, July/August
1966, pp. 274-280.
6.0 AC Motor Speed Control F~om DC Supply
6.1 The Through-Pass Inverter and its Application to the Speed
Control of Wound Rotor Induction Machines, P. N. Miljanie, IEEE Transactions on Power Apparatus and Systems,
Vol. PAS-87, No.1, January 1968, pp. 234-239.
6.2 Induction Motor Speed Control with Static Inverter in
Rotor, A. Lavi, R. Polge, IEEE Transactions, Vol. PAS-85,
No.1, pp. 76-84.
6.3 Modulating Inverter System for Variable-Speed Induction
Motor Drive (GM Electrovair II), R. W. Johnson, IEEE
404
CHOPPERS, INVERTERS AND CYCLOCONVERTERS
6.4
6.5
6.6
6.7
6.8
6.9
6.10
6.11
6.12
6.13
6.14
6.15
6.16
6.17
6.18
6.19
6.20
Transactions on Power Apparatus and System, Vol. P AS-88,
No.2, February 1969, pp. 81-85.
A Wide-Range Static Inverter Suitable for AC Induction
Motor Drives, P. M. Espelage, J. A. Chiera, F. G. Turnbull,
IEEE Transactions of the IGA, Vol. IGA-5, No.4, July
1969, pp. 438-445.
New Inverter Supplies for High Horsepower Drives, R. P.
Veres, IEEE/IGA Conference Record, October 1969, pp.
537-545.
PWM Inverters for AC Motor Drives, B. Mokrystzki, 1966
International Convention Record, Part i, pp. 8-23.
Optimum Design of an Input Commutated Inverter for AC
Motor Control, S. B. Dewan, D. L. Duff, IEEE/IGA Conference Record, 1968, pp. 443-455.
Precise Speed Control With Inverters, A. J. Humphrey,
IEEE Conference on Industrial Power Conversion.
SCR Inverter for Deep Submergence Propulsion Systems,
R. Gilbert, J. Langton, Conference Proceedings SAE Aerospace Systems Conference, July 1967, p. 17.
"Controlled Power-Angle Synchronous Motor Inverter
Drive System," G. R. SIemon, J. B. Forsythe and S. B.
Dewan, IEEE/IGA Conference Record 1970, pp. 663-667.
"Method of Multiple Reference Frames Applied to the
Analysis of a Rectifier Inverter Induction Motor Drive,"
P. C. Krause, J. R. Hake, IEEE Transactions paper 1969
Winter Power Meeting.
"The Controlled Slip Static Inverter Drive," B. Makrytzki,
IEEE/IGA Transaction, May/June 1968, pp. 312-317.
"Variable Speed Induction Motor Drive System for Industrial Applications," R. W. Johnston, W. J. Newill, IEEE/
IGA Conference Record, October 1970, pp. 581-585.
"Slip Power Recovery in an Induction Motor by the Use of
a Thyristor Inverter," W. Shepherd, J. Stanway, IEEE
Transactions on IGA, Vol. IGA-5 #1, January/February
1969, pp. 74-83.
"Thyristor DC Switch Inverter," T. Kume and R. G. Hoft,
IEEE/IGA Conference Record 1971, pp. 299-312.
"Current Source Converter for AC Motor Drive," Ken P.
Philips, IEEE/IGA Conference Record 1971, pp. 385-392.
"An Inverter for Traction Applications," J. T. Salihi, IEEE/
IGA Conference Record 1971, pp. 393-400.
"Characteristics and Applications of Current Source/Slip
Regulated AC Induction Motor Drives," Robert B. Maag,
lEEE/IGA Conference Record 1971, pp. 411-416.
"AC Commutatorless and Brushless Motor," T. Maeno and
M. Kobato, IEEE/IGA Conference Record 1971, pp. 25-34.
"Wide Speed Range Inverter," E. F. Chandler and F. N.
Peters III, IEEE Transactions on Industry and General
Applications, Vol. IGA-6 #1, January/February 1970, pp.
19-23.
405
SCR MANUAL
'~Several
Modulation Techniques PWM Inverter," R. D.
Adams and R. S. Fox, IEEE/IGA Conference Record 1970,
pp. 687-693.
6.22 "Fundamentals of a Pulse Width Modulated Power Circuit,"
L. J. Penkowski and K. E. Pruzinsky, IEEE/IGA Conference
Record 1970, pp. 669-678.
6.23 "A Pulse Width Modulated Three Phase Complementary
Commutated Inverter" by S. B. Dewan and J. B. Forsythe,
IEEE/IGA Conference Record 1971, pp. 321-326.
6.24 "Harmonic Analysis of a Synchronized Pulse Width Modulated Three Phase Inverter," J. B. Forsythe and S. B.
Dewan, IEEE/IGA Conference Record 1971, pp. 327-332.
6.25 "The Programmed Bridge Regulator: A New Approach to
Efficient Power Conversion," E. H. Philips and R. D. Underwood, IEEE/IGA Conference Record 1971, pp. 333-339.
6.26 "A Wide Speed Range Inverter Fed Induction Motor
Drive," R. L. Rigberg, IEEE/IGA Conference Record 1969,
pp. 629-633.
6.27 "An Investigation of an SCR Inverter Drive for an Induction
Motor, D. M. Mitchell and C. J. Triska, IEEE/IGA Conference Record, October 1967, pp. 81-90.
6.28 "The Through Pass Inverter and Its Application to the
Speed Control of Wound Rotor Induction Machines," P. N.
Miljanic, IEEE Transactions on Power Apparatus & Systems, Vol. PAS-87 #1, January 1968, pp. 234-239.
7.0 Pulse Modulator Switches
7.1 High Power Thyristor-Battery. Drive for High Peak, Low
Average Power Pulser, V. Wolk, Proceedings of IEEE,
Special Issue on High-Power Semiconductor Devices, Vol.
55, No.8, August 1967, p. 1454.
7.2 Magnetic Pulsers from PAM Multiplexed Instrumentation,
K. Aaland, G. A. Pence, Workshop on Applied Magneties,
May 1969.
.
7.3 Pulse Generators, Glasoe & LeBacqz, McGraw-Hill Book
Co., Inc., New York, 1948.
7.4 A 300 KW Semiconductor Magnetron Modulator, F. A.
Gateka and M. L. Embree, 1962 International Solid-State
Circuits Conference, University of Pennsylvania, February
16,1962.
7.5 How to Get More Power From SCR Radar Modulators, T.
Hamburger, C. H. Wood, R. A. Gardenghi, Electronic
Design, September 13, 1963.
7.6 Some Characteristics of Thyristors in High-Power Modulator Circuits, T. H. Robinson, Modulator Symposium, May
1966.
7.7 Adding SCR's to get High Power Means Smaller Transmitters, C. R. Brainard, W. R. Olson and E. H. Hooper, Electronics, June 13, 1966, pp. 119-126.
8.0 Induction Heating, Ultrasonics, Lighting
8.1 A Static Power Supply for Induction Heating, J. P. Landis,
IEEE Transactions on Industrial Electronics and Control
6.21
406
CHOPPERS, INVERTERS AND CYClOCONVERTERS
8.2
8.3
8.4
8.5
8.6
8.7
8.8
8.9
8.10
8.11
8.12
8.13
8.14
8.15
8.16
8.17
Instrumentation, Vol. IECI-17, No.4, June 1970, pp.
313-320.
An SCR Inverter With Good Regulation and Sinewave Output, N. Mapham, IEEE/IGA, Vol. IGA-3, No.2, March
1967, pp. 176-187.
Thyristor Power Units for Induction Heating and Melting,
R. S. Segsworth, S. B. Dewan, IEEE/IGA Conference
Record, October 1967, pp. 617-620.
An Ultrasonic Power Source Utilizing a Solid-State Switching Device, W. C. Fry, IRE International Convention
Record, Part 6, Vol. 8, 1961, pp. 213-218.
An SCR Inverter With Good Regulation and Sinewave
Output, N. Mapham, Application Note 660.16,* General
Electric Company, Syracuse, New York.
Thyristor Control of Fluorescent Lighting Banks, J. L. Storr,
lEE Conference, No. 17, Power Applications of Controllable Semiconductor Devices, November 1965, pp. 178-185.
A Low Cost, Ultrasonic-Frequency Inverter Using a Single
SCR, N. Mapham, Application Note 200.49, * General Electric Company, Syracuse, New York.
Dimming Fluorescent Lamps, J. C. Moerkens, Philips Technical Review, Vol. 27, No. 9/10, 1966, pp. 265-273.
A Solid-State Supply for Induction Heating and Melting,
S. B. Dewan and G. Havas, IEEE/IGA Conference Record,
November 1969, p. 686.
Power Thyristor High Frequency Limits, R. L. Davies,
IEEE International Conference Record, March 1969, p. 198.
"A 180 KW 8-11 kHz Thyristor Frequency Converter for
Induction Heating," Ivan Horvat, IEEE/IGA Conference
Record 1971, pp. 837-849.
"Practical Design Considerations for Regulated Sine Wave
Inverter," Walter B. Guggi, IEEE/IGA Conference Record
1971, pp. 869-876.
"Latest Developments in Static High Frequency Power
Sources for Induction Heating," B. R. Pelly, IEEE Transactions on Industrial Electronics and Control Instrumentation, Vol. IECI-17 #4, June 1970, pp. 297-312.
"A High Frequency Power Supply for Induction Heating
and Melting," G. Hauas and R. A. Sommer, IEEE Transactions on Industrial Electronics and Control Instrumentation, Vol. IECI-17, No.4, June 1970, pp. 321-326.
"A Static Power Supply for Induction Heating," J. P. Landis,
IEEE Transactions on Industrial Electronics and Control
Instrumentation, Vol. IECI-17, No.4, June 1970, pp.
313-320.
"A Solid State Supply for Induction Heating and Melting,"
S. B. Dewan and G. Hauas, IEEE Transactions on IGA,
NovemberlDecember 1969, pp. 696-692.
"AC to AC Frequency Converters for Induction Heating
and Melting," S. B. Dewan and G. Havas, IEE Proceedings
Conference on Power Thyristors and Their Applications,
London, May 6-8, 1969, pp. 440-447.
407
seR
MANUAl
8.18 "New Developments in High-Frequency Power Sources,"
W. E. Frank, IEEE Transactions on IGA, Vol. IGA-6 #1,
January/February 1970, pp. 29-35.
8.19 "Oscillator Circuit Thyristor Converters for Induction Heating," E. Golde and G. Lehman, Proceedings of the IEEE
Special Issue on High Power Semiconductor Devices, Vol.
55, #8, August 1967, pp. 1449-1453.
8.20 "Power Supply Systems for Induction Furnaces," R. S. Segsworth and S. B. Dewan, IEEE Conference Record of the
IGA 1970, pp. 279-283.
8.21 "Thyristor Power Units for Induction Heating and Melting,"
R. S. Segsworth and S. B. Dewan, IEEEIIGA Conference
Record, October 1967, pp. 617-620.
8.22 "Sine Wave Inverter System," Walter Guggi, IEEE/IGA
Conference Record 1970, pp. 517-524.
408
LIGHT ACTIVATED THYRISTOR APPLICATIONS
14
LIGHT ACTIVATED THYRISTOR APPLICATIONS
Light, or more precisely, electromagnetic radiation, with wave
length between 0.2 and 1.4 microns, is increasingly being used in conjunction with solid state devices. The use of light offers a convenient
method for sensing the absence or presence of an opaque object and
for achievmg electrical isolation. These features are useful in the control
of power devices and will be discussed in this chapter. Opto-electronic
devices also find usage in communications, and other applications
beyond the scope of the Manual.
14.1 LIGHT ACTIVATED SEMICONDUCTORS
There are many types of devices available for converting radiant
energy into electrical information. These devices may convert the
radiant information into a variable resistance, such as occurs in photo
resistive elements such as cadmium sulfhide, cadmium selenide and
lead sulfhide, or into a generated voltage and current as in photo voltaic
cells of selenium, silicon and germanium. The newer types of light
sensitive devices are semiconductor junction devices. Included in this
group are light activated diodes, transistors, and thyristors. It is to this
last group of devices which we will direct our attention.
Radiant energy incident on a semiconductor (such as silicon,
germanium, cadmium, sulfhide or selenium) causes the generation of
hole-electron pairs. These free charges create a change in the electrical
characteristics of the semiconductor. In a photo cell they cause a
decrease in resistance, in the photo-voltaic cell they create a voltage
and in a junction device, under bias, cause currents to How across the
exposed junctions. In a photo thyristor this current is. equivalent to
gate current, whereas in a photo transistor the light causes an equivalent base current.
14.1.1 Photo Diode (Light Sensitive Diode)
In a conventional p-n junction without external bias, a very thin
depletion region is formed where electrons from the n-type material
move across the junction and combine with holes in the p-type material. The positive ionss(} created in the n-type material and the negative
ions in the p-type material build up an electric field.
409
SCR MANUAL
~ fP-TYPE) .
FYPE
IN-T~YPE
•
:i
DISTANCE
(I) Without Bias
FIGURE 14.1
(b) Reverse Bllsell
PH JUNCTION AND CARRIER CONCENTRATION
In a reverse biased p-n junction, the width of this depletion region
will increase proportionally with applied voltage (capacitance will
decrease with higher voltage). Electrons which attempt to enter the
p-type material are too low in energy to .cross the potential barrier and
the current will be almost zero. Some electrons will be excited by
thermal energy to an extent that hole electron pairs are created which
will be swept across the junction as leakage current by the existing
field of the depletion region.
If light (electromagnetic radiation) with the proper wave length
is directed toward the reverse biased p-n junction, absorbed photon
energy will also create 'hole electron pairs and enable the electrons to
move across the depletion region. -Additional current proportional to the
light intensity will flow across the reverse biased junction.
+1p
VA
r
-
,
-1p
-
+
(a) Hole Electron Pairs in Reverse Biased
L1pt Sensitive PN Junction
+VA
"0
",
".
"3
-1p
(b) VI Characteristic of Lilllt Sensitive
Reverse Biased PN Junction
FIGURE 14.2
- Ip
'1J
q
= '1J • q . View of Pboto Transistor
(b> Symbol
IG
H.40mW/cm 2 ..........
\y././
./
4!1
"
MAXIMUM POWER / '
DIS~TlON
/
40
H = 20mW/cm~"""'"
/'
./
/,/"
.,
..'"IE
/
35
\/
'"
:>
u
...:I:
30
/
H.lo,\w/cm;//'
/
25
co
:J
,
/'
/
I
I
E
/
20
-='
15
10
5
1"/
il \
///
/
/
/
A
H-5mW/cm 2
)..././
V X
V
/
/'
./
;/
/
~
TUNGSTEN BULB
2870· K
-----
H'2m~m2_
10
20
30
40
50
veE - COLLECTOR TO EMITTER VOLTAGE - VOLTS
(e) VI Curves Vs Light Intensity for the
L14A502 Photo Transistor
FIGURE 14.4
Switching speed is an important characteristic of this device and
should be considered. For the L14A502 the delay time is about 2 p..sec
and rise time about 5 p..sec.
Note: Response time is heavily dependent on external base-to-emitter
impedance since collector-base capacitance is multiplied by
Miller effect. Most significant in darlington!
412
LIGHT ACTIVATED THYRISTOR APPLICATIONS
LIGHT
td
=
DELAY TIME
= RISE TIME
" = STORAGE TIME
tr
If
= FALL
TIME
(OPEN BASE)
td
FIGURE 14.5 RELATIONSHIP BETWEEN INPUT • OUTPUT OF LIGHT SENSITIVE DEVICE
14.1.3 Photo Darlington Amplifier
Like the photo transistor, the current between collector and emitter
of the photo darlington amplifier is a function of the light incident on
the device. The dominant term on the output current is the product of
the two betas, which accounts for the high sensitivity of the device.
N
p
la) Simplified P.."ical Layaut of PIlato
Darllngtan Amplifier
(II) Photo Darllngten Amplifier lIIustratlnl the
Effects of PIIotDn CUrrant GanaratlH
FIGURE 14.8
lEI = Ip1 (hFEI + 1)
IE2 (IP2 + lEI) (hFE2 + 1)
IE2
[IP2 + IP1 (hFEI + Ipl)] [hFE2 + 1]
Because Ip2 is small compared to lEI:
IE2 "'" Ip1 . hFEI • hFE2
IE
Emitter Current
Ip = Photon Produced Current
hFE = DC Current Gain of Transistors 1 and 2
The 2N5777-2N5780 family is one. example of a light sensitive
darlington. Its spectral response is centered near 0.85 microns, maximum collector current is 250 mA, power dissipation 200 mW at 25°C
and rise time is typically 75 ,...seconds. These devices can be used with
or without base lead. Base bias can increase or decrease sensitivity
depending on the bias polarity.
=
=
=
413
SCR MANUAL
~
~
--
(a) View of Darlington Amplifier
~
l.--
!...
o.1
.~
20
'0_
3-
~
2
I-- ~
J.0
I...
-
I-- I - '--""I - -
~
.
~
e
40
-- --- r
0
mW/cm2
---- -
100
~
1. - -
,...
0.5
l..-- ~
~
0.2
l..--- f--
-
~
NORMALIZED TO:
VCE=5V
H** = 2mWI cm 2
,
'0
'5
20
.0
25
.5
veE-COLLECTOR TO EMITTER VOLTAGE-VOLTS
E
**H-RADIATION FLUX DENSITY. RADIATION SOURCE IS AN UNfiLTERED TUNGSTEN
FILAMENT BULB AT 2870· K COLOR TEMPERATURE.
(tI) Symbol
(c) Normalized light Current Vs
Collector to Emitter Voltage
FIGURE 14.7
LIGHT SENSITIVE DARLINGTON AMPLIFIER
14.1.4 Light Activated SCR (LASCR)
The basic operation of a light activated SCR is shown in Figure
14.8. With applied forward voltage junctions Jl and Is are forward
biased and they can conduct if sufficient free charge is present. Junction J2 is reverse biased, however, and blocks current flow. Light entering the silicon creates free hole-electron pairs in the vicinity of the J2
depletion. region which are then .swept across J2 (Note: The theory
developed for the photo diode and extended to the transistor can be
applied here also). As light is increased the current in the reverse
biased diode, Figure 14.8(c}, will increase. The current gains of the
n-p-n and p-n-p transistor equivalents in the structure also increase
"LIGHT"
\\
J,
OEPmgioN
A
REGION
P
.-.-~
.~
.~
C
(a) & (b) Simplified Physical layout of LASeR
FIGURE 14.8
414
(c) LASCR Transistor Equivalent (d) Symbol of LASCR
Illustrating the Effects of
Photon Current Generation
& .Juncti on Capacitance
LIGHT ACTIVATED SCR
LIGHT ACTIVATED THYRISTOR APPLICATIONS
with current. At some point the net current gain (al + a2) exceeds unity
and the SCR will turn on. The criterion for turn-on is the same as
explained in Chapter 1 but with an additional term due to the light
generated current.
I _ a2 (Ip ± IG) + IcBo (1) + IcBo (2)
A - I - a2 - al
=
=
=
Ip
Photon Current (Current Generated by Incident Light)
IG
Gate Current
ICBO(l) + I cBo (2) = Leakage Current
a
Current Gain
a]
varies with IA
+ (Ip)
+
a2 varies with IA
(Ip ± I G)
when al
a2 ~ 1 than IA ~ 00
+
In order to obtain reasonable sensitivity to light the SCR must be
constructed so that it can be triggered with a very low current density.
This requires the use of a fairly thin silicon pellet of small dimensions
hence high current devices are not considered practical for light triggering at this time. The high sensitivity of the LASCR also causes it to
respond to other effects which produce internal currents. As a result
the LASCR has a higher sensitivity to temperature, applied voltage,
rate of change of applied voltage, and has a longer turn-off time than
a normal SCR.
Some important characteristics of the L8-L9 series light activated
SCR's. are shown in Figure 14.9(a)-(e)..
NOTES, (I) IRRADIATION FROM TUNGSTEN SOURCE.
(2)
CURVE DEPICTS TYPICAL VARIATION OF TRIGGERING
SENSITIVITY WITH ANGLE OF IRRADIATION.
JUNCTION
TEMPERATURE-~
(8) Light Triggering Characteristics
FIGURE 14.9
(b) Typical Angular Response
CHARACTERISTICS OF THE L8·L9 LASCR
415
8CR MANUAL
v
1.0
~E:
0.8
""'\
/
CURVE DEPICTS RELATIVE RESPONSE OF
THE UGHT ACTIVATEO SCR AS A FUNCTION
OF THE WAVELENGTH OF THE INCIOENT
ENERGY.
\
V
/
\
\
/
f--
0.2
SCATTERED LIGHT FROM
HOUSING ON EDGE OF ~
PELLET.
-
-
V
V
L~
"
.,'v:
/
\
\\
'-fOLLIMATED LIGHT
ONTOP OF PELLET
ONLY.
/'
o
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
WAVELENGTH-MICRONS
(e) Typical Spectral Response
1.0
0.9
0.8
.........
........
r---..
0.7
r- r-
;--.
0.6
r- roo-
g
Ii
0.5
iii:
-
r-- '"- r--
NOTE: THIS CURVE DEFINES THE RATIO OF EFFECTIVE IRRADIANCE
TO TRIGGER WITH ANY APPUED ANODE VOLTAGE TO THE
0.4 f-- f-EFFECTIVE IRRADIANCE TO TRIGGER WITH 6 VOLTS APPLIED
ANODE VOLTAGE.
-
l - t--
0.3
0.2
O. I
40
60
80
120
100
160
140
INS.TANTANEOUS APPLIED ANODE VOLTAGE-VOLTS
(d) Typical Variation of Uglrt sensitivitJ Willi Anode Voltage
10
9
8
7
6
NOTE: THIS CURVE DEFINES THE RATIO OF
EFFECTIVE IRRADIANCE TO TRIGGER
WITH A SPECIFIC RESISlOR FROM
GATE TO CATHODE TO THE
EFFECTIVE IRRAIllANCE TO
TRIGGER WITH 56,000 OHMS
FROM GATE TO CATHODE.
I......
5
4
~
"- ........
3
"- ""I'..
f'4000
6000
10000
8000
20000
40000 60000
GATE TO CATHODE RESISTANCE-OHMS
(e) l'JDlcal Variation of Ught sensitivity Willi
Gate to cathode Resistance
FIGURE 14.9 CHARACTERISTICS OF THE U-LI LASCR
416
180
200
LIGHT ACTIVATED THYRISTOR APPLICATIONS
Figure 14.9(a) shows the relationship between effective irradiance
to trigger and junction temperature. Since the triggering level is highest
at the lowest junction temperature, the amount of light provided in a
given system must take into account the lowest junction temperature
at which operation is expected. Conversely the maximum sensitivity
is obtained at maximum junction temperature. Therefore, the maximum irradiance provided by the system under conditions where the
LASCR should not trigger must be less than the value given for
the highest operation junction temperature.
Figure 14.9(b) shows the spatial response typical of the L8 and L9
devices. Primary response is confined to an angle of ±25° from perpendicular. Secondary responses are found at about 55 0 from the
perpendicular. Secondary responses are produced by high reHection
from the inside walls of the case. The specified effective irradiance to
trigger is given for a point source of light oriented perpendicular to the
plane of the silicon pellet.
Figure 14.9(c) shows the relative response of the LASCR as a
function of wave length of the incident irradiation. This curve also
indicates the difference in spectral response of the LASCR to direct
light and scattered light. Direct radiation must penetrate a significant
thickness of silicon to reach the region of J2. Since the absorption of
silicon is rather high in the visible spectrum (0.4-0.7 micron) the spectral
response is rather low in this band. Scattered light from the housing
reaches the vicinity of J2 near the edge of the pellet, hence less of the
shorter wave length is absorbed before reaching the junction. The result
of edge radiation is higher response to the shorter wave lengths.
Figure 14.9(d) shows the typical variation of light sensitivity with
anode voltage. At high voltages, the required irradiance to trigger
becomes significantly less as a result of the effect of voltage on the
gain of the equivalent transistor circuit. Since HET is normally specified
with an anode voltage of 6 volts, operation at higher or lower voltages
will modify this value. If the applied voltage is sinusoidal and the
irradiance is increased slowly from low level, triggering will occur
initially at the peak of the applied wave form. Further increase in irradiance will then advance the point of triggering to the beginning of
the applied wave.
Figure 14.9(e) shows that typical light sensitivity is inversely proportional to gate to cathode resistance. The purpose of the gate cathode
resistor is to bypass current around J1, thus reducing the gain of the
n-p-n transistor region to desensitize the device. The use of temperature sensitive resistors (thermistors) between gate and cathode or a
forward biased silicon diode plus resistor network can provide some
degree of temperature compensation against changes in sensitivity. It
should be noted in Figure 14.9(a) however that the effect of temperature upon sensitivity is far from consistent from one device to the
next. Therefore, it is not practical as a general rule to provide temperature compensation which will maintain light sensitivity constant
over the operating temperature range.
The General Electric LASCR is similar to the C5 type planar SCR
except that there is a glass window on top of the package. The device
is capable of handling up to 1.6 amperes RMS anode current and of
blocking up to 200 volts peak.
417
SCR MANUAL
CATHODE
ELECTRODE
'"
LEAD
,
APPLIED
'I FllRWARO
J2 VO+~E
LIGHT
"I
GATE
... NEGATIVE
J3 POSITIVE
HERMETIC
SEAL
(II) LASCR Planar Pellet
WEUlED
\~
.,,~
SILICON PELLET
MAIN SEAL
LIGHT SENSITIVE
AREA
CATHODE
(e) LASeR Symbol
(a) LASCR Construction
FIGURE 14.10 CONSTRUCTION OF HOUSING, PELLET Ie SYMBOL OF LASCR
14.1.5 Light Activated Silicon Controlled Switch (LASCS)
The light activated silicon controlled switch (LASCS) is another
planar thyristor structure with all four semiconductor regions accessible,
rather than only three as is customary with silicon controlled rectifiers.
Accessibility of the fourth region greatly expands circuit possibilities
beyond those of conventional transistors or SCR's. In addition, making
it sensitive to light adds an entirely new dimension to the circuit design
possibilities. It is probably the most versatile p-n-p-n device on the
market today.
CATHODE
GATE
ANOO~E4
2 'CATHODE(O)
p
p
p
N
N 3 ANODE
GATE
(II) LASCS Planar Pellet
~ ANODE
(a) View of LASCS
:f
"
CATHODE
ANOIlE
GATE
GATE
CATHODE
(e) LASCS Symbol
FIGURE 14.11
HOUSING, PELLET & SYMBOL OF LASCS
The theory developed for the LASCR can be employed for the
LASCS. Tum-on and turn-off techniques used for the SCR and LASCR
are applicable to the LASCS, but the anode gate gives the added possibility of tum-on with negative pulses with respect to the anode and
tum-off with positive pulses.
418
LIGHT ACTIVATED THYRISTOR APPLICATIONS
ANODE
CATHODE
FIGURE 14.12
LAses
TRANSISTOR EQUIVALENT
The General Electric light activated silicon controlled switch is
electrically similar to the 3N80 series silicon controlled switch, but
is provided with a lens capped package for triggering by light (see
Figure 14.11{a». The curves shown for the LASCR are similar for the
LASCS.
14.2 LIGHT EMITTING DEVICES
14.2.1 Tungsten Lamps
Tungsten filament incandescent lamps are probably the best
known light emitting devices. Their wide range of spectral emission,
their good efficiency and low cost make them ideal devices to use with
General Electric light sensing devices .
...
...
,
.I
'.4
...
......,,,,,,,.-V
"
t
i;
V
,
V
V
V
V
V
V
v
,./
'.0
[7
V
vV'
V
V
V
..
VEFFECTI\IEIlCINLIIISCRl
RADIANT EMlnMCE
V
10-
1/
V
.,
.I
M00280028CJOlIOOO~~
.'
g
3
V
18
V
'l(
'0
V
1.2
V
14
16
FII,..AMENT
(a) Emittance of Tungsten
FIGURE 14.13
Ie
2.0
Z"
1£ MPERATURE. T ,oK
24
l(
26
i?8
l.O
32
34
1000 TUNGSTEN LAMP
(b) Relationship Between Color Temperature and
True Filament Temperature for Tungsten
Lamps
MOST IMPORTANT CHARACTERISTICS OF TUNGSTEN LAMPS
419
SCR MANUAL
..00
I
~400
"
I
..
/
LAIilPS
-
I
I-
I
I
I
I
I
/
VACUUM LAMPS
l-
I
I
RAIt8U
I
I
I
!
/
.,,,nu",
I
i
,
1I
I
~ NOfn.AL
~-
V
/
I-" "
....
/
P ... OTOFLOOO
V
I
/
II
/
+-
.
•G
o
,
,
20
"HIRe MeT-TO-COLD MSISTANCE II'"TtO
(c) Color Temperatura Vs Hot·to·Cold
Resistance Ratio for Tungsten
...
I
TUNGSTEN LAMPS
!l\."~
1000
.... .......
V
f\\..\."G.:
G~?"
2IlOO
z
~
........
V
noD
yl-""
!:!
0:
co
,.".,/
2400
:!:
~.;s
/
I-""
~r
2200
/
2000
1800
o
.2
.8
.6
.4
1.0
1.2
1.4
1.6
1.8
MSCPJWATT
{MEAN
SPHERICAL.
CANOLEPOWER
PER WATT 1
FIGURE 3.9 APPROX. COLOR TEMPERATURE
vs
EFFICIENCY
(d) Color Temperatura Vs Efficiency
FIGURE 14.13
420
MOST IMPORTANT CHARACTERISTICS OF TUNGSTEN LAMPS
LIGHT ACTIVATED THYRISTOR APPLICATIONS
y
50
2.7S
V~
2.!10
t ....
i
20
~
...
;
..
1;
I."IS
1.110
1.25
--
>
Ii
r
1.0
.1!1
..... ~
.80
.25
o
r - r--
~EfFECTIVE
RADIATED fUJX
,,
-
~
o
/
-7 ~
10-
i/
1/
_
DENSITY
/
-
~~R.T\!""
~
/'
.".-/
v
60
40
100
80
-
180
140
120
(e) Effects of Voltage
on Tungsten Lamps
II
.-~-+--+--+--t-~--t-~-4--4--+--+--+--+--4
101-~-4--4--+--+--+--t-~--t-~-4r-4--+--+-~
:J
~81--+--t---+--t--+--t--+-+--+-+--+-4--+-~--I
tlj
Z7H--+-t---+-t--+-+--+-+--+-+--+-4--+-~--I
..
~61t:t-+-t---+-t--+-t--+-+--t-+---t-4---t-~--I
a
E51H-+-t--+-+--+-+--+-+--+-+--+-4--+-~-4
0:
g
4
If~fj~~\l-:.t---j--t---+-+--+-+--+-+--+-4--+-4--Jf.--J
{l~
(f)
M:::~:.................. .
Inrus" Current
NORMAL~
Vs Time With
Rated Voltage
Applied
CUR~ENT
I
o
10
20
30
40
50
60
70
80
90
100
110
120
MILLISECONDS
80
80
70
90
IOOOX
!IOOX
::;
c
Iz
.!II
....
E
:J
110
130
"\.
IOOX
/
r-...
v
f'{.-......
20X
"'~
.","
"-
lOX
.,..
5X
"'- "-
.x
(l)C/laraGtlr
Clll'Yes of
Tungsten Lamps
120
140300
\
-
200X
~
60
~ 'T
==-CURt:I--""/
_~fO"'E"
_ c~..fI>.E 1
70
I
80
./
...,eV
V
......~
CURRENT
"' " r-...
90
PERCENT OF DESIGN VOLTS
.".-
110
~/ v
-=
I
I
__ I
25
100
IX
5X
2X
~ ""'120
140
V
.1 X
;;I
i
19
..
05XE
I"100
-
130
ISO
TIME (mS)
100
130
......
140
OIX
FIGURE 14.13 MOST IMPORTANT CHARACTERISTICS OF TUNGSTEN LAMPS
421
SCR MANUAL
Figure 14.13(a) shows the high radiant emittance of tungsten as
a function of color temperature and also its high effective radiant
emittance on the LASCR. Effective irradiance to trigger a LASCR
typically is between 0.15 mW/cm2 at 100°C to 20 mW /cm2 at -65°C.
This indicates that any tungsten lamp will trigger an LASCR providing
the distance is not excessive. The same can be said for all other GE
light sensing devices mentioned in this chapter.
Color temperature (CT) is a very popular measurement for tungsten lamps because color temperature defines the majority of the lamp's
characteristics. Several of these characteristics are shown in Figure
14.13(a)-(b).
One very useful method of establishing filament temperature is
based upon the fact that resistance of the filament increases with
increasing temperature. Figure 14.13(c) shows the theoretical relationship between color temperature and the ratio of hot resistance to cold
resistance, measured at 25°C. The measurement of this cold resistance
must be made with the lowest voltage and current possible in order
to prevent heating during the measurement. This can be done with a
bridge by applying the voltage momentarily and noting the initial direction of the galvanometer deflection, then balancing accordingly. The
observed resistance should be decreased by about 10% (multiply by
0.9) to account for resistance of lead wires, socket, and the connecting
ends of the filament which are cooled by the lead wires. Measurement
of hot resistance is made by using operating voltage and current values.
Small lamps having low-mass filaments will have considerable cyclic
temperature variations when operated at 50 or 60 Hz, hence hot resistance should be measured with an oscilloscope.
Figure 14.13(d) shows an easy method of approximating color
temperature by the luminous efficiency of the lamp. If the input power
and either the mean spherical candlepower (MSCP), or candlepower
(CP) or total output lumens (I
F/4on-) (Source intensity
1 in
lumens/steradian = candle; Total flux output of source = F in lumens)
are known, then the color temperature may be established. The difference between evacuated and gas-filled lamps is the result of heat being
conducted away from the filament by the gas. In general, lamps
designed for operation at 5 volts or less, and less tpan 10 watts, are
evacuated. It should also be noted that the geometric configuration of
the filament will produce variations from the data of Figure 14.. 13(d).
Once a point has been established for a lamp, it may be desirable
to know the effect of variations in supply voltage on the effective radiant
output. Figure 14.13(e) shows this effect upon output and upon color
temperature.
Similar information is given in Figure 14.13(f) which also relates
life, current, and candlepower to lamp voltage. Note that the life curve
is on a logarithmic scale. At 65% voltage, life is extended 200 times,
input power is 50 %, and effective radiation on an LASCR is reduced
to 40% of the initial value.
Of particular importance is the effect of "normal" supply voltage
variations of ±10%.
When incandescent lamps are connected in series with· semiconductors the initial lamp inrush current flows through the semiconductor.
=
422
=
LIGHT ACTIVATED THYRISTOR APPLICATIONS
Peak inrush current can be up to 20 times the normal RMS operating
current. It will not always reach the maximum values shown in Figure
14.13(g) because circuit impedance will have a limiting effect. Applying a preheat voltage is often used to limit inrush current to safe values.
See Section 9.2.2 for further discussion on this subject.
14.2.2 Light Emitting Diodes (LED) or Solid State Lamps (SSl)
The light emitting diode is a p-n junction which when forward
biased will emit light. There are several inherent advantages of an SSL
light source over conventional sources.
1. Very fast response time. Fall and rise time can be in the order
of a few nanoseconds to a few microseconds (depending on
type).
2. Long life and mechanical ruggedness, leading to much improved reliability.
3. Low impedance of SSL, similar to a conventional forward
biased diode.
4. The predominant light output is monochromatic.
Light emission .from a p-n junction occurs when electrons from
the bottom of the conduction band recombine with holes at the top
of the valence band. The energy released from the electron corresponds
to the width of the forbidden energy gap.
r - - - - - - - - - , - - - - --
~f]
-.....:
t
'-----------'
~ HEAT RELEASE
LIGHT RELEASE
- - -
+
_
E
---
(a) Energy Bands In SSL
(b) Biasing Of SSL
FIGURE 14.14 ENERGY BANDS AND BIASING OF SSL
A gallium arsenide lamp has an energy band gap of EG
The wave length which will be emitted (,\) is:
h· C.
.
lI. = ~ III mIcrons
= 1.37 eV.
Q
where:
h
C
EQ
For CaAs: EQ
= 6.63 X 10- joule seconds (Planck's constant)
= 3 X 10 micron/second (velocity of light)
34
14
= Energy in Joules (lev = 1.6 X 10- 19 joule)
= (1.39) (1.6 X 10- 19) = 2.19 X 10- 19 joule
_ (6.63 X 10- 34 ) (3 X 1014 ) _ 905 . 10- 1 .
,\ 2.19 X 10-19
-.
mICrons
.905 microns
This calculated wave length belongs to the infrared region of the
radiant energy spectrum. CaAs SSL therefore are classified as infrared
sources since they have a peak emission near 0.9 micron.
The energy released by the electrons, which may occur as light
or heat, has to be replaced by an external power source as shown in
=
423
SCR MANUAL
Figure 14. 14(b). The quantum efficiency (QE) of a SSL determines
how much electrical energy is converted to light energy. In an indirect
material, some of the electrons when traveling from the conduction
band to the valence band are detained in trapping levels, causing
phonon (heat) as well as photon (light) releases.
QE = No. of Photons Out
No. of Electrons In
(b) Symbol
(a) View of SSL
FIGURE 14.15 VIEW OF SSL AND SYMBOL OF SSL
f-(/)
~
::::;
{!/
f--
6
lI-
5
_ 300
j'11
f--
I-
Z
~ 250
II
TC :2S·C
..J
!
f--
~ 200
4
ir
7
3
~ 150
o
::>
7
o
...
o
I
II:
~
'\
I-
I-
ir
I-
\
...
II:
I
I
V
o
t1J
N
5\._4
1/
.>
'"
.4
.6
50
-
~
I
.2
-
::::;
'"
II:
.8
1.0
1.2
,
SSL-4
....... :P ....
100
SSL-5A
SSL-58
SSL-5C
~
-75 -50 -25
0
25
100
e:,'b....: fA
~
Il
II
!.;
!.;
40
\-,,<0,.-
~~..,
PULSE WIDTH I ,II SEC.
RE~ RATE 200 PPS
TEMPERATURE IS 2S"C
I
I
60 80 100 120 140 160 180 200 220
?OWER OUTPUT 1M ILLIWATTS)
(e) Power Output Vs Peak
Current
FIGURE 14.16
424
50
75
---..
100 125· 150
(b) Case Temperature Vs
Normalized Power Output
(a> Forward Current Vs
Power Output
20
.-.::.: "
CASE TEMPERATURE I"C)
FORWARD CURRENT I AMPS CONTINUOUS)
I~o
~
I
o
1.4
IF"IOO rnA
LIGHT ACTIVATED THYRISTOR APPLICATIONS
The most useful information in an SSL specification sheet is its
power output- (irradiance) H out in mW versus current input. Figure
14.16(a} shows the power output H in mW increasing until a peak
point is reached. The efficiency of GaAs SSL decreases with increasing
temperature because more trapping levels occur in the energy gap
(see Figure 14.16(b) ).
To increase output power Hout. SSL's are often used in a pulsed
condition. SSL's can withstand much higher peak currents under these
conditions. The average current depends on the duty .cycle which is
determined by: Duty Cycle
= ton +ton.
Maximum peak current Ip
toff
is dependent on maximum average current and the duty cycle:
I
Duty Cycle
I avg •
=
peak
14.3 PHOTON COUPLER
Light emitting devices and light sensing devices have major applications in areas where electrical isolation between the input signal and
the output is important.
FIGURE 14.17 CIRCUIT COMPONENTS IN A PHOTO COUPLER
The General Electric PC15-26 and PC4-73 photon couplers consists of an SSL and a photo transistor. When an input signal is applied, ..
to the GaAs SSL, the light emitted is detected by the photo transistor .
and converted back to an electrical signal. Input and output signal have
complete electrical isolation and there is no feedback from the output
to the input.
The main advantages are:
1. Simple interface between different voltage levels.
2. Noise isolation and ground loop elimination.
3. High speed switching.
4. Modification of output signal through access to base
terminal.
5. High shock and vibration immunity.
6. Bounceless switching.
7. Small size.
Figure 14.18 shows input current versus output current for the
PC4-73. Note that this is a linear relationship. Output current and
dark current are influenced by temperature and should he considered
when designing with photon couplers.
425
SCR MANUAL
240
;(
.5
!z
200
/
~ 160
~ 120
o
t;
/
80
OJ
'"<.>
H
40
VCE=IOV
/
:::>
OJ
I-
V
V
TA =25·C
/
0102030405060708090
IF-SSlINPUT CURRENT (mAl
FIGURE 14.18
INPUT CURRENT (mAl VS OUTPUT CURRENT (mAl FOR THE PC4·73
PHOTON COUPLER
14.3.1 Specifications of Light Intensity*
In order to apply these devices, it is necessary to know if a given
source at a known distance will cause the desired response in a sensor.
This is a function of the characteristics and intensity of the light, the
response of sensor to that· type of light, and the physical relationship
and optical coupling between the source and sensor. The output of
most sources is specified in terms of visible light. There is no general
relationship between visible light and the effect of the radiation on a
sensor.
·see application Note 200.34 for more detailed discussion of this subject and applications of
the LASCR.
14.4 CHARACTERISTICS OF SOURCES AND SENSORS
Most people learn in high school physics that light is a form of
electromagnetic radiation. Electromagnetic radiation is characterized
by its frequency (or more commonly in the case of light, wave length),
magnitude and direction. Sources vary greatly in the components of
frequency in their output. Figure 14.9(a) shows the spectral distribution of some of the more common types of light sources. The characteristics of the light from a tungsten lamp are a function of the color
temperature of that lamp. The color temperature depends upon the
type of lamp and upon the applied voltage.
426
LIGHT ACTIVATED THYRISTOR APPLICATIONS
Ii(j;.
C
ULTRAVIOLET I I
-~"!~~~
BLUE
GREEN
YELLOW
ORANGE
RED
INFRARED
--
100%
eo
60
DARLINGTON AMPl.
40
20
o
0.2
0.4
0.6
O.B
1.0
1.2
1.4
1.6
1.8
2.0
1.8
2.0
A.-WAVELENGTH-MICRONS
(a) Ught Sensitive Devices
to 100%
.
..""
;!;
80
L
0
0
...
N
..
:;
60
40
C
IE
20
0
Z
0
0.2
0.4
0.6
O.S
1.0
1.2
1.4
1.6
).-WAVELENGTH-MICRONS
(b) Light Emitting Devices
FIGURE 14.19 SPECTRAL DISTRIBUTIONS OF GENERAL ELECTRIC LIGHT SENSITIVE AND
LIGHT EMlmNG DEVICES
The effect of electromagnetic radiation on a sensor depends upon
the wave length of the radiation. The relative effect of radiation of
different wave lengths upon the eye and upon several types of silicon
semiconductor sensors is shown in Figure 14.9(b). The eye responds
to shorter wave lengths than do silicon devices. By comparing Figure
14.9(a) and 14.9(b) it can be seen that most of the radiation emitted
by a tungsten lamp is not visible. Hence, it is important to note that
the amount of visible light produced by a source does not reveal how
effective this source will be upon a silicon sensor.
14.4.1 Definition of Light Intensity
The intensity of electromagnetic radiation incident on a surface is
called irradiance (H). Its dimensions are watts/square centimeter. Since
any type of electromagnetic radiation has a spectral distribution, it is
also reasonable to define the irradiance per unit of wave length (HA).
HA is a function of the wave length. By definition then,
H=fHAdA
YA is the relative response of a sensor to electromagnetic radiation
at any given wave length. The effect of a particular wave length from
a light source on a given sensor is the product HA YA. The effect of
radiation on a sensor is additive so that to determine the total effect
427
SCR MANUAL
of a particular source on a particular sensor,.jt is necessary to add the
H,\ Y,\ products for all wave lengths of interest. Or,
HE = fIlA Y,\ and capacitor C 2,
charges to approximately 200 volts through R2 and Ra. When the
master flashgun fires (triggered by the flash contacts on the camera)
its light output triggers LASCRh which then discharges capacitor C2
into the primary winding of transformer T l' Its secondary puts out a
high voltage pulse to trigger the flashtube. The flashtube discharges
capacitor C h while the resonant action between C 2 and T 1 reverse
biases LASCR1 for positive tum-off. With the intense instantaneous
light energy available from present-day electronic flash units, the speed
of response of the LASCR is easily in the low microsecond region, leading to perfect synchronization between master and slave.
High levels of ambient light can also trigger the LASCR when a
resistor is used between gate and cathode. Although this resistance
could be made adjustable to compensate for ambient light, the best
solution is to use an inductance (at least one henry) which will appear
as a low impedance to ambient light and as a very high impedance
to a flash.
'T
I
UTC NO. PF7
I
I
I
I
I
I
I
300VDC
SOURCE
CI
TI
+
IOOO,.F
+
C2
o.22,.F
..L
G-E
1t3
1.8M
LASCRI
L8B
FT-I06
FIGURE 14.31 SlAVE FlASH
441
SCR MANUAL
14.5.12 Ught Activated Motor Control
Light sensing devices can be used to perform the switching function in a reversible motor control circuit. In this case SI has to be used
.
to reset the circuit.
Figure 14.40 shows such a light activated control where the light
is used to control the direction of rotation of a balanced winding permanent split capacitor motor through two triacs.
r- ------------,
I
115 VAC
OR
220VAC
50Hz
OR
60Hz
I
I
:I
CR,
iL
I
MOTOR
:I
v~
R.
__________ J
I
I
LASCR: GE L9U
CRII CR2 - GE AI4F
FIGURE 14.40 REVERSING INDUCTION·MOTOR DRIVE
The transformer T 1 is selected to have a dc secondary voltage VI
between 6 to 24 volts.
R -R -~-~-~
R5=10ohms
1 2 - IGatel - IGate2 -100ma
Whenever light is directed toward LASCR!> triac 1 would be triggered, turning the motor into one direction. Removing the light from
LASCR1 and directing it toward LASCR2 would turn off triac 1 and
turn on triac 2 which will reverse the direction of the motor.
LASCR1 and LASCR2 could be replaced by LASCS or by light
sensitive transistors but maximum current rating for the transistor
should be considered. The triacs must have voltage ratings at least
equal to the capacitor voltage rating (usually 1.5 to 2.5 times peak
line voltage).
14.6 CIRCUITS FOR LIGHT EMITTING DEVICES
Designing power supplies for light emitting devices requires the
understanding of how the wave shapes of the supply influence the
irradiance, H, of light emitting devices.
*Measurement was made with the A14 diode In series.
442
LIGHT ACTIVATED THYRISTOR APPLICATIONS
WAVEFORM
00
INCANDESCENT
LAMP
LIGHT EMITTING
DIODE
HOUT [mW/CM2]
HOUT [mW/CM2]
180 0
o LINE
AC
o LINE
DC 1/2 WAVE
I
o LINE
__- L_ _
~
__
~
__
~~
_ _L -_ _
~
__
~
_ _- L_ _
~
__
~~
_ _L -_ _
L-~
DC PULSES
*MEASUREMENT WAS MADE WITH THE AI4 DIODE IN SERIES
FIGURE 14.41
DEPENDENCE OF IRRADIANCE OF LIGHT EMITTING DEVICES FROM
INPUT WAVEFORMS
The irradiance H in the above table was measured on a GE #1813
light bulb (14.4 volts; 0.1 amp; = 0.86 CP). b.H maximum will be
larger for smaller, lower mass filament bulbs and will be smaller for
larger bulbs. If b.H, the irradiance change as shown in the table,
cannot be tolerated, filtered dc supplies should be used.
14.6.1 Low Loss Brightness Control
A circuit which changes average value of the DC supply voltage
on the light emitting device is shown in Figure 14.42. Because of the
high switching frequency the tungsten lamp will have an almost continuous adjustable light output between 0 and 100%. If a light emitting diode is used as the emitting device, the irradiance will be in phase
with the applied current pulses and will decrease to zero when the
supply current is zero.
443
SCR MANUAL
R,
GE NO.-4~4G
4.7V .5A
'OK
6V
R.
'8K
ALL RESISTORS' 1/2W
FIGURE 14.42 LOW· LOSS BRIGHTNESS CONTROL
In this circuit the PUT is used as an oscillator; the time constant
and resultant frequency are determined by (R3 + R4) X C 1 • Every time
Ql fires, Q2 will be forward biased, driving Qa into saturation and
applying the battery voltage to the light bulb.
14.6.2 Current Limiting Circuits
To protect light emitting devices from damaging current levels,
different types of limiting circuits are used.
-I,
+
0,
~~
J'sc
FIGURE 14.43 SIMPLE CURRENT LIMITER
A simple form of current limiter is shown in Figure 14.43. At low
currents Ql is forward biased applying the full power supply voltage
to the load. When 11 . Rl "'" VOR!> Ql will come out of saturation and
limit lamp current to this value.
A more effective current limiter is shown in Figure 14.44. When
the voltage drop across Rl is greater than the base threshold voltage
of Q2, Q2 will begin to conduct, divert base drive from Ql and Ql will
limit the output current.
+
M
"f
SSl
FIGURE 14.44 HIGHER PERFORMANCE CURRENT LIMITER
444
LIGHT ACTIVATED THYRISTOR APPLICATIONS
14.6.3 Impulse Circuits for Light Emitters
Because solid state light emitters have very low emission levels
when continuously forward biased, a very popular method is to use
them in a pulsed mode. Current levels can be many times higher than
the continuous current without exceeding the average power.
A low current pulser can be very easily built with unijunction
transistors of the same type used for SCR triggering circuits. More
information on these circuits can be found in Chapter 4.
v
v
v
+
+
+
R,
0,
R3
C,
Ca)
.,
••
R.
R,
PUlser With Unljunctlen
Translstar
R.
rn....mmable
Unljunctlen Translster
(b) Pulser With CemplementarJ Ce) Pulser With
UnlJunctien Translster
FIGURE 14.45 UNIJUNCTION TRANSISTOR PULSE GENERATORS
Pulse circuits for higher current levels can be designed by using
SCR's which are ideal devices for such applications.
+
o.SA LOW INDUCTANCE
FIGURE 14.46 HIGH CURRENT PULSE GENERATOR
In Figure 14.46 capacitor C 1 is charged through Rl and discharged through SCR1 when the SCR is triggered. The limiting factor
of this circuit is the holding current of SCR1 which determines the
smallest value for Rb
E1max
RlmlD
=-1-Hmin
445
SCR MANUAL
Because the recharge cycle is long (five times Rl . C 1) this circuit is
limited to low frequencies if large capacitors are used. Much higher
frequencies can be obtained when a fast charging path for C 1 is added
to the above circuit as in Figure 14.47. Rl supplies Ql with base drive,
+------,-------,
lis
0.5 OHM LOW INDUCTANCE
FIGURE 14.47
HIGH CURRENT HIGH FREQUENCY PULSE GENERATOR
turning Ql on. Current into the capacitor is limited by Rs. After C 1 is
charged and SCR 1 is fired, CR1 is forward biased and as long as CR1 is
conducting, the base of Ql is at a lower potential than the emitter.
Assuming Vs/Rl < holding current of SCRb SCR1 will turn off and the
cycle will repeat.
REFERENCES
1. The Light Activated SCR, E.K. Howell, Application Note 200.34,
General Electric Company, Syracuse, N. Y.*
2. Series Operation of Silicon Controlled Rectifiers, J. C. Hey, Application Note 200.40, General Electric Company, Syracuse, N. Y.*
3. How to Use the New Low Cost Light Sensitive Transistor - The
LI4B, Neville Mapham, Seminar Note 671.8, General Electric
Company, Syracuse, N. Y.*
4. Solid-State Optoelectronics in '69, R. E. Koeper, Electronic Design
News, February 1969.
5. A Course on Opto Electronics, Jack Hickey, et al, The Electronic
Engineer, July, August, September 1970.
6. Simple Circuit Gives Fast, High Current Pulses to Drive a GaAa
Laser Pulser, J. R. Frattarola, Ideas For Design, Electronic Design,
December 1967.
7. Physics of Semiconductor Devices, S. M. Sze, Wiley Interscience
1969, pp. 625-730.
8. Optical Engineering Handbook, J. A. Mamo, Editor, General
Electric Company, Ordnance Department, Pittsfield, Mass.
9. Solid State Lamp Manual, Solid State Lamps - Part I (3-8270)
and Part II (3-0121), General Electric Company, Nela Park, Cleveland, Ohio.
10. Flashtube Data Manual, Photo Lamp Department #281, General
Electric Company, Nela Park, Cleveland, Ohio.
*Refer to Chapter 23 for availability and ordering information.
446
PROTECTING THE THYRISTOR AGAINST OVERLOADS AND FAULTS
15
PROTECTING THE THYRISTOR AGAINST
OVERLOADS AND FAULTS
Satisfactory operation of thyristor circuits and the equipment in
which they operate often depends heavily on the ability of the system
to survive unusual overcurrent conditions. One obvious answer to this
requirement, not usually an economical one, although one that is
becoming more reasonable with the decreasing costs of semiconductors, is to design the system to withstand the worst fault currents on a
steady-state basis. This requires semiconductors and associated components that are rated many times the normal load requirements.
Where this approach is not possible because of economics or other
factors, an adequate overcurrent protective system is usually used.
15.1 WHY PROTECTION?
The functions of an overcurrent protective system are any or all
of the following:
1. To limit the duration of overloads and the frequency of application of overloads.
2. To limit the duration and magnitude of short circuits.
3. To limit the duration and magnitude of fault current due to
shorted semiconductor cells. 1
The objective of these functions is to safeguard not only semiconductor components but also the associated electrical devices and
buswork in the equipment from excessive heating and magnetic
stresses. The trend toward high capacity systems feeding electronic
converter equipment often results in extremely high available fault
currents. Since both heating and magnetic stresses in linear circuit
elements respond to the square of the current, the importance of adequate protection in "stiff" systems is self-evident.
Elaborating on function No.3 above, thyristors as well as diode
rectifiers may fail by shorting rather than by opening. In many circuits
such a device fault results in a direct short from line to line through
the low forward resistance of the good devices in adjacent legs during
+
SHORTED
CELL
FIGURE 15.1
ARROWS INDICATE FLOW OF FAULT CURRENT THROUGH GOOD DEVICE AFTER
ADJACENT LEG HAS SHORTED. LOAD RESISTANCE DOES NOT LIMIT CURRENT
447
SCR MANUAL
at least part of the cycle, as illustrated in Figure 15.1. Under these
circumstances, a protective system functions either to shut the entire
supply down or to isolate the shorted device in order to permit continuity of operation. This will be discussed at greater length later.
It is difficult to make broad recommendations for overcurrent protection since the concept of satisfactory operation means diHerent
levels of reliability in different applications. The selection of a protective system should be based on such individual factors as:
1. The degree of system reliability expected.
2. The need or lack of need for continuity of operation if a semiconductor fails.
3. Whether or not good semiconductor cells are expendable in
the event of a fault.
4. The possibility of load faults.
5. The magnitude and rate of rise of available fault current.
Depending upon the application, these various factors will carry
more or less weight. As the investment in semiconductors increases for
a specific piece of equipment, or as an increasing number of components in a circuit increases the possibility of a single failure, or as continuity of operation becomes more essential, more elaborate protective
systems are justified. On the other hand, in a low-cost circuit where
continuity of operation is not absolutely essential, economy type semiconductors may be considered expendable and a branch circuit fuse in
the AC line may be all that is needed, or justified, for isolation of the
circuit on faults, allowing semiconductor components to fail during
the interval until the protection functions. In other designs, the most
practical and economical solution may lie in overdesigning the current
carrying capability of the semiconductors so that conventional fuses
or circuit breakers will protect the semiconductors against such faults.
It is therefore· reasonable that each circuit designer rather than
the semiconductor component manufacturer decide precisely what
level of protection is required for a specified circuit. Once the specific
requirements are determined the component manufacturer can recommend means of attaining these specific objectives. This chapter is
prepared to assist the circuit designer in determining his protection
requirements, and then to select satisfactory means of meeting these
requirements.
15.2 OVERCURRENT PROTECTIVE ELEMENTS
The main protective elements can be divided into two general
classes. One class consists of those devices which protect by interrupting or preventing current How, and the other class consists· of those
elements which limit the magnitude or rate of rise of current How by
virtue of their impedance.
Among the elements in the first class are:
1. The AC circuit breaker or fuse which disconnects the entire
circuit from the supply.
2. The cell fuse or breaker which .isolates faulted semiconductor
cells.
3. Load breakers or fuses which isolate load faults from the equipment or a faulted cell from DC feedback from the load or
. parallel converter equipmerits.
448
PROTECTING THE THYRISTOR AGAINST OVERLOADS AND FAULTS
4. Current limiting fuses and SCR circuit breakers.
5. Gate blocking of SCR's to interrupt overcurrent.
Internal leads of stud mounted devices may burn open under
severe fault currents before regular protective elements function. Prior
to leads burning open, the associated junction will have been permanently damaged to a shorted condition. (The internal fusing characteristics of commercial semiconductors are generally not predictable nor
reliable enough to be used as protective elements in practical circuits.)
Press Pak packages most generally fail short.
Among the elements of the second class which limit magnitude
or rate-of-rise of current are:
1. Source impedance.
2. Transformer impedance.
3. Inductance and resistance of the load circuit.
15.3 CO-ORDINATION OF PROTECTIVE ELEMENTS
Depending upon their complexity and the degree of protection
desired, converter circuits include one or more of the various interrupting devices listed above. Functioning of these devices must be coordinated with the semiconductor and with each other so that the overall
protection objectives are met. Fuses or breakers must interrupt fault
currents before semiconductor cells are destroyed. In isolating defective semiconductors from the rest of the equipment, only the fuse or
breaker in series with a defective semiconductor cell should open.
Other fuses and breakers in the circuit should remain unaffected. On
the other hand, when a load fault occurs, main breakers or fuses should
function before any of the semiconductor cell-isolating fuses or breakers
function. This fault discriminating action is often referred to as selectivity. In addition, the voltage surges developed across semiconductors
during operation of protective devices should not exceed the transient
reverse voltage rating of these devices. More complex protective systems require meeting additional coordinating criteria. The example
of a protection system and its associated coordination chart described
later in this discussion illustrates some of the basic principles of coordination for both overloads and stiff short circuits.
The magnitude and waveshape of fault and overload currents
vary with the circuit configuration, the type of fault, and the size and
location of circuit impedances. Fault currents under various conditions
can generally be estimated by analytical means. References 1, 2, and 3
show analytical methods for calculating fault currents for generally
encountered rectifier circuits.
For overloads on rectifier or inverter circuits where the current is
limited to a value which the semiconductors can withstand for roughly
50 milliseconds, conventional circuit interrupting devices like circqit
breakers and fuses can usually be used satisfactorily for protection.
This type of overload can be expected where a sizeable filter choke
in the load or a "weak" line limits the magnitude or rate of rise of
current significantly or where semiconductor components are substantially oversized. By placing the circuit breaker or fuse in the line ahead
of the semiconductors, the protective device can be designed to isolate
449
SCR MANUAL
the entire circuit from the supply source whenever the line current
exceeds a predetermined level which approaches the maximum rating
of the semiconductors for that duration of fault.
For time intervals greater than approximately 0.001 second after
application of a repetitive overload, the thyristor rating for coordination purposes is determined by the methods discussed in Section 3.6.
If the overload being considered is of a type that is expected only
rarely (no more than 100 times in the life of the equipment), additional
semiconductor rating for overload intervals of one second· and less can
be secured by use of the surge curve and J2t rating for the specific
device being considered.
The surge characteristic is expressed as the peak value of a halfsine wave of current versus the number of cycles that the semiconductor can handle this surge concurrent with its maximum voltage, current,
and junction temperature ratings. In circuits that do not impose a
half-sine wave of fault current on the semiconductors, the surge curve
can be converted into current values that represent the particular waveshape being encountered. 4 The surge curve for the semiconductor can
be converted to different waveshapes or different frequencies in an
approximate, yet conservative, manner for this time range by maintaining equipment RMS values of current for a specific time interval.
For example, the peak half-sine wave surge current rating of the C35
SCR for 10 cycles on a 60 Hz base is shown on the spec sheet to be
88 amperes. For a half-sine waveshape, the RMS value of current over
the complete cycle is one-half the peak value,. or 44 amperes. To convert this to average cell current in a three-phase bridge feeding an
inductive load (120-degree conduction angle), divide this RMS value
by y'&(44 -;- y'3 = 25.4 amps). To determine the total load current
rating for a bridge using this cell, multiply the average cell current
by 3. (25.4 X 3 = 76.2 amps).
15.4 PROTECTING CIRCUITS OPERATING ON STIFF POWER
SYSTEMS
Conventional circuit breakers and fuses can be designed to provide adequate protection when fault currents are limited by circuit
impedance to values within the semiconductor ratings up to the time
when these protective devices can function. However, circuits requiring
good voltage regulation or high efficiency will usually not tolerate
high enough values of series impedance to limit fault currents to such
low values unless substantially oversized semiconductors are used.
When a fault occurs in a circuit without current limiting impedance,
current will develop in a shape similar to the dashed line in Figure
15.2. Its rate of rise is limited by the inductance inherent in even the
stiffest practical systems. If the peak available current substantially
exceeds the semiconductor ratings, and if it is permitted to How in the
circuit, the semiconductor would be destroyed before the current
reaches this first peak. Conventional circuit breakers and fuses will
not function quickly enough. Instead, "current limiting" fuses which
melt extremely fast at high levels of current are used. Alternately,
"electronic circuit breakers" of the type discussed in Section 8.8 can
be designed for this purpose.
450
PROTECTING THE THYRISTOR AGAINST OVERLOADS AND FAULTS
PEAK
_ _ _ AVAILABLE
/'
i
....z
or
'"
~
o
/
/
/
I
"CURRENT
\
/
®
PEAK LET-THRU
-CURRENT
\
\
\
\
\
\
TlME--+
FIGURE 15.2
LIMITING ACTION OF CURRENT LIMITING FUSE
15.4.1 The Current Limiting Fuse
The terms "current-limiting" fuse applies to a fuse which, when
used within its current-limiting ratings, limits the peak let-through.
current to a value lower than that which would overwise How in the
circuit. The action of a typical current limiting fuse is indicated in
Figure 15.2. Melting of the fuse occurs at point A. Depending on the
fuse design and the circuit, the current may continue to rise somewhat
further to point B, the peak let-through current. Beyond this point the
impedance of the arcing fuse forces the fault current down to zero at
some point C.
The interruption ratings become very important when applying
fuses to power systems having high fault current capability. A fuse with
an interruptive rating less than the fault capability of the system at the
location of the fuse may not be able to interrupt a short circuit within
its clearing I 2 t and peak let-through rating, resulting in damage to
the power semiconductor it is protecting.
It becomes obvious that time and current are the controlling
factors in the function of fuses, and the time-current let-through characteristics of the fuse must conform to the time-current rating of the
SCR the fuse is protecting. References 4, 5, 6 and 7 discuss in detail
the behavior and characteristics of current limiting fuses based on
specified conditions of physical surroundings, circuit parameters and
fuse design. A practical method of coordinating fuse characteristics to
that of SCR's under specified conditions 'i·8 is discussed in the following
section.
15.4.2 Fuse·SCR Coordination in AC Circuit
Before setting down a set of logical design steps let's look at a
typical fuse-SCR circuit (Figure 15.3) in order to define terms and
become acquainted with typical waveshapes. Assume a fault across
451
SCR MANUAL
SCRI
FI
,--- Z L ----'\
SCR2
(
RL
I
LL
I
I
Zs
l
THYRECTOR
ASSUME0:>k
FAULT
I
I
I
'" VSOURCE
I
FIGURE 15.3
'1
ZLO
J
TYPICAL FUSE·SCR CIRCUIT
the load impedance ZLQ. In Figure 15.4 this fault is shown _taking
place close to the instant of peak source voltage. This is the most
stringent condition for the fuse to interrupt under large prospective
fault current conditions and high circuit X/R ratios.
YSOURCE-
i SCRI
--
Time
1000Amps 10010-
Occu,enCf
ISCR!
100010010I FUSE
1--
FIGURE 15.4
CIRCUIT WAVEFORM OF FIGURE 15.3 UNDER STEADY STATE &
TYPICAL TRANSIENT FAULT CONDITIONS
X/R ratio as used here refers to the ratio of the series reactive
to resistive elements of the circuit when shorted. Since the series
reactive component in power circuits is nearly always inductive, the
ratio of X/R is a relative measure of the energy a circuit can store.
Since this energy must be absorbed by the fuse in its arcing phase
it serves as an indication of the severity of the current quenching duty
placed upon the fuse. Figure 15.5 shows a close-up of the fuse action
shown in Figure 15.4.
452
PROTECTING THE THYRISTOR AGAINST OVERLOADS AND FAULTS
-~-VSOURCE
VSOURCE
/
"
I~
RMS
I
Tim,
I,
I FUSE
al SCRI
r - - - -.. Time
FIGURE 15.5 CIRCUIT WAVEFORMS DURING FUSE CLEARING INTERVAL
The fuse current wavefonn is typically triangular in shape with
an effective pulse width of te seconds and a peak of t amperes. It is
important to note that t" can vary from less than 1h ms to greater than
8 ms in 60 Hz circuits, while the variation in t is typically from 10 to
100 times fuse RMS rating. t and t" are the parameters that detennine
the destructive effects of the short circuit current, both on the semiconductor and on other circuit components.
15.4.2.1 Fuse Ratings
Fuse manufacturers generally give only the following data:
• Values of J2 t at different RMS circuit voltages and prospective
fault currents
• Peak let-through current curves vs RMS prospective current
• Melting time vs RMS current curves
The latter curve's time values shouldn't be confused with 4 since
the values given for melting time rarely extend below 10 ms and the
time values are for melting time only - not complete fuse clearing
time. Thus, these curves are not very useful for short circuit current
evaluation of fuse behavior. They are valuable only for long tenn fuse
overload conditions.
Figures 15.6 and 15.7 show typical fuse perfonnance in a 480 V
circuit. These two curves when taken together characterize the fuse
as a function of the circuit parameters, VSOURCE and Ip. For a given
J p , tc can be found by solving Equation 15.3.
3 (J2 t)
tc=~
Both t and t" as a function of circuit parameters can be found from
the fuse manufacturer's data.
453
SCR MANUAL
..
ft
U
...'"...
E
c
30A
-
. . IOO~~m±!jjfftl~±ItiAj
,:
Ip A.ailaili. Current lA IS,mm. RIIS)
FIGURE 15.8 FUSE PERFORMANCE IN 480 V RMS CIRCUIT
J
ii
~y
.J.
~,
_ ~.t.i'
- -..:.~ ....
.
-~,.
..,.
,....,.
::::: p
.... i-'
40A
:SOA
.... i-'
I6A
;.-
;...... ... ........
'& ....,
o
........
0.1
FIGURE 15.7
1.0
10
100
%p. Avallalll, Curr.'" lA IS,m... RMS)
FUSE PERFORMANCE IN A 480 V RMS CIRCUIT
Figure 15.9 is provided to aid in the conversion of data from the
format consisting of Figures 15.6 and 15.7 to 15.8 which provides a
common basis of comparing SCR capability with fuse let-through
performance.
454
PROTECTING THE THYRISTOR AGAINST OVERLOADS AND FAULTS
/0
6
Nominal Fuse Ra'ln,
Ampi RMS
4
4~00 50
\
~
2
~
c
!:
"
35
:~O.,/ ~'"
"1&,,("[/K"
~~R~
I
u
... 0.8
o
IL 0.6
..
"7
Circuit Prolpective Curren'
20
Ip~mm"'lcall
~5
. .......:
""-=
.' ..... ......... ..,
4.2
2
IX
~
I
~
~
1/
0.4
"'V
~
boo.
17 ~ 'Y
1/
17""
0.2
.5
~
~
.•.
j
17
0. I
0.1
0.2
0.4
0.6 0.8 I
2
Clearina Time-ml
4
6
8 10
FIGURE 15.8 FUSE-5CR APPLICATION CHART. THE FUSE PORTION OF THE CHART IS OERIVED
FROM FIGS. 15.6 AND 15.7 AND THE NOMOGRAPH SHOWN IN FIG. 9. TABLE 1
SHOWS THE DERIVATION PROCEDURE FOR THE 16 A FUSE LINE
I
Kilo Am,.
20.
15
10.
8
FIGURE 15.9 FUSE CLEARING
TIME NOMOGRAPH
12,
KIlO Amp2_ ••c.
'c
6
5
10.0.
Mlili leconds
10.
8.0
6.0.
2
1.5
1.0.
0..1
D.'
0.5
0..2
Dol
0..4
E....'II:
'c'!lrla
0.002
J • 3.31C1l0 AM,.
Itead:
0..0.01
1 2 , • ,0. Kilo A.,2_ ..c•
0..3
0..2
0..5
0..1
455
SCR MANUAL
00
800
··..
... 600 u
~:.\~y'c'l'P ~::-:I
01
E
400
...c 30
..,.:,. ~I"'"
•'!' 250
..
.;
---
IC,7
200
L150
.·["iI
0
E
c,
FIIURE 15.10 I.. AND I VS PULSE
WIDTH FOR A 35 A RMS SCR
~
~
600
....
,..
~
!i 400
u
~
300
250
i 200
150
i
L100
1.5 2 2.5 3
4
6
8 10
itu's. 1_ Width·ml
Fuse circuit definitions
tm Fuse melting time
tA Fuse arcing time
tc
tm + tA fuse dearing time
1
Peak instantaneous fuse let-through current
1 = y2
VA
if
Vsource
(15.1)
Z.+~
Maximum symmetrical rms circuit fault current
Abbreviated prospective current
Peak fuse arc voltage
Instantaneous fuse current
12 t c
=
f
to
+ tc
i2 f dt
= clearing J2t
(15.2)
to
= 12 t m
+ 12t A
= (12/3) te for a triangular waveform
Note: I as used in 12tc is a rms current value.
For a triangular waveform:
12 tc
(15.3a)
12t =
12 t c
(15.3b)
1 = _}
2 (12 t)
12 t =
3
And for a half sinusoidal waveform:
'1
2
*
tc
*sinusoidal waveform 12 t value
456
(15.4)
Available fault current symmetrical RMS
(Prospective current Ip) in kA
1) From Figure 15.6
2) From Figure 15.7
3) Data from 1, 2 above
using Figure 15.9
J2t (A 2 s)
II (kA)
1:" (ms)
0.5
52
0.21
3.5
1
55
0.265
2.5
2
60
0.34
1.6
5
65
0.47
0.9
10
70
0.6
0.58
20
77
0.79
0.37
50
85
1.1
0.21
100
90
1.3
0.16
=:"
:::0
Fuse data conversion from manufacturer's data sheet to fuse-SCR
application chart. 480 V circuit voltage~ 16 A fuse.
~
z
G')
TABLE I
-i
:::I:
m
-i
A list of fuse manufacturers supplying current limiting fuses is
given in Table II.
Chase-Shawmut Company
General Electric Company
Power Systems Management
347 Merrimac Street
Newburyport, Massachusetts
Department
6901 Elmwood Avenue
01950
Philadelphia, Pennsylvania 19142
English Electric Corporation
Bussmann Manufacturing Division One Park Avenue
McGraw-Edison Company
New York, New York 10016
St. Louis, Missouri 63100
Carbone-Ferraz Inc.
P. O. Box 324 (Elm Street)
Rockaway, New Jersey 07866
"'......"
(,J'I
TABLE II
MANUFACTURERS OF CURRENT LIMITING FUSES
:::I:
-<
:::0
~
:::0
~
z>~
~
:::0
5><::I
en
>z
<::I
~
c:
Si
SCR MANUAL
15.4.2.2 SCR Rating For Fuse Application
SCR rating data for use with fuses is provided by means of subcycle surge curves as discussed in Section 3.5.5 .and shown in Figures
3.8 and 15.10. Note the curves are provided for haH sinusoidal pulse
waveshapes for testing convenience. Since the fuse let-through current
waveshape is triangular the designer must account for the difference
in waveshape upon SCR surge capability. From test results and
analytical studies it has been shown that matching SCR peak current
capability with that of the fuse let-through capability provides a conservative basis for SCR protection by means of current limiting fuses.
Use of 12t provides a gross error if matched directly. That is: an SCR
can typically withstand a current surge having a sinusoidal waveshape
J2t value that is 150% higher than that of a triangular current surge
waveshape of the same pulse base width.
15.4.2.3 Selecting aFuse For SCR Protection
Fuse selection is simplified by plotting SCR sub-cycle peak current
capability (obtained directly from the SCR data sheet), directly over a
family of fuse characteristics (for the given circuit voltage conditions) as
shown in Figure 15.8; [formerly the RMS current was given. (If using
an old data sheet convert RMS to peak current prior to plotting.)]
A fuse current rating is then selected such that the SCR rating curve
exceeds that of the fuse let-through current rating curve under worst
case circuit prospective current conditions.
The reason for showing the semiconductor curves dotted below
1 ms results from a lack of vigorous test data on the semiconductor's
short term surge capability in the 100 to 1000,us range. It is known
that due to dil dt restrictions, peak current capability of an SCR is
reduced below approximately 100/Ls pulse width. Until firm test data
is available, use discretion in this range.
Worked Example
Referring to Figure 15.3 assume
1. The source transformer is rated 100 KVA with 5% short circuit impedance.
2. To select a fuse/SCR combination in a 480 volt RMS circuit
with a 20 ampere RMS load. The fuse must be able to protect
the SCR in case of a fault occurring across the load.
Design Steps
1. Choose the SCR. Based upon voltage and current considerations a C137 is chosen with a case temperature of 90°C at
9 amperes average per SCR to deliver a 20 ampere RMS load
current.
2. Obtain Figure 15.8 information from fuse manufacturer or plot
from· manufacturer's data of the form shown in Figures 15.6
and 15.7.
3. Superimpose on it the C137 sub-cycle surge current data
derived from SCR data sheet.
4. Calculate the maximum RMS symmetrical available short circuit (prospective) current assuming that the transformer react458
PROTECTING THE THYRISTOR AGAINST OVERLOADS AND FAOLTS
ance is the only available short circuit impedance
IA =
KVA
P.U. Impedance X KV
100
(.05) (,480)
~ 4.2 KA
5. Select the fuse. The fuse current rating must be greater than
the load current but must limit the maximum let-through current to a value below the sub-cycle surge current capability
of the SCR. In this case a fuse rating of 25 amperes will properly protect the SCR.
6. Check fuse manufacturer's data to determine fuse arcing voltage. In .many circuits it should not exceed the SCR's rated
voltage as SCR's adjacent to the device being protected may
be called upon to block the fuse arcing voltage.
7. To avoid nuisance fuse blowing mount the fuse such that the
SCR does not contribute to fuse heating. Provide thermal isolation by providing adequate distance between SCR connections
and that of the fuse.
15.4.3 Fuse·SCR Coordination in DC Circuits
The fundamental problem in applying fuses for the protection of
SCR's in either AC or DC circuits is to ensure that the fault let-through
energy comes within the withstand capability of the SCR under all
circumstances which can arise in service. Protection can only be
applied to the extent that the essential parameters and conditions to
be met can be identIfied and specified in the properly related terms.
In the case of AC applications, the parameters on which the SCR
withstand capability is normally compared to the fuse are:
1. Peak let-through current (and thus available current) versus
clearing time
2. Clearing I 2t (in absence of #1)
3. Applied voltage
4. Power factor
In the case of DC applications, the essential parameters become:
1. Peak let-through current versus clearing time
2. Clearing I 2 t (in absence of #1)
3. Applied voltage
4. Rate of rise of fault current, di/dt
5. Time constant
The fuse interrupting behavior in AC and DC circuits is essentially different, consequently there is simply 0110 relationship between
AC and DC fuse performance. It is therefore necessary for fuse manufacturers to supply separate DC peak let-through current values in
relation with fuse clearing time in order to coordinate with the SCR's
sub-cycle capability. Figures 15.11 and 15.12 show one manufacturer's
method of relating energy, current and time coordination in terms of
fuse design and circuit parameters. Based on these figures a fuse-SCR
coordination curve,Figure 15.13, can be easily plotted. The DC fuseSCR coordination curve can be interpreted and applied in a similar
way as that of the AC curve discussed in the previous section.
459
SCR MANUAL
TIME CURRENT CHARACTERISTIC AND I2T CURVES
1316
NOTES:
I. 1100 VOLTS DC
"
104
If)
2. LlR -9OMS
3. 8 AMPS TO 63 AMPS
10 3
e
A
I
-
10 2
'"o
l!i
\
I0
~
10
AVAILABLE CURRENT
I-
lI:
...
IIOOV
100
lA) IN MULTIPLES
0
.1
'OF CURRENT RATING (IN)
'"
1!!
....
I
;:
e'
I
-..
2
nOOVDC
~
3
I
10
100
RMS MELTING CURRENT AND AVAILABLE CURRENT (IA)
IN MULTIPLES OF CURRENT RATING UN}
FIGURE 15.11
TIME CURRENT CHARACTERISTIC AND 12t CURVES
MAXIMUM PEAK LET THROUGH CURRENT CHARACTERISTIC CURVES
, NOT':".:
1/
'2. LlF ',90, M,S •.
/
IZ
II!0:
::>
u
y
'"co
~
/
'"
......
"~
lI-
.
.i
i
1/
/
•
~§
III
~~~
-
lO
AVAILABLE CURRENT UA)(KA}
FIGURE 15.12
460
MAXIMUM PEAK LET THROUGH CURRENT CHARACTERISTIC CURVES
PROTECTING THE THYRISTOR AGAINST OVERLOADS AND FAULTS
800
AVAILABLE FAULT CURRENT
~~~~I..J
NOTES
1000VDC RATED
CURVE PLOTED AT "OOVDC
50A RATED
40A RATED
200r-----1-~~---
"f---"'I"-...:::f----lf--l7'9""';'- 30A RATED
150t-----+----+-------l-----t-:r""t---+-t---lM-2OA RATED
.~
AVAILABLE SYMMETRICAL
FAULT CURRENT
IOOI'';:O------:!15,---~2;t;O,-------::=---!:::-----;:5='=O----;:6='=O~70;:-:!60~90::cI~OO:
CLEARING TIME -MILLISECONDS
FIGURE 15.13
DC FUSE-5CR COORDINATION CURVE
15.5 INTERRUPTED SERVICE TYPE FAULT PROTECTION WITHOUT
CURRENT LIMITING IMPEDANCE
By inserting the protective device in the AC lines feeding a semiconductor AC to DC converter, protection can be provided both against
DC faults and semiconductor device faults if there is no possibility of
DC feed into faults of the semiconductor devices themselves. DC feed
into cell faults will occur in single-way circuits when other power
source:; feed the same DC bus or when the load consists of CEMF
types of loads such as motors, capacitors, or batteries. The following
is an example of this type of AC line protection in a circuit without
current limiting impedance. Upon functiOning of the protective system,
the circuit is interrupted and shut down.
Worked example: Referring to Figure 15.T5
Assume:
- 120 V RMS AC supply, 60 Hz.
- Single-phase bridge employing two C35H SCR's
for phase control in two legs and two IN2156 diode
rectifiers in the other two legs. See Figure 15.14.
- Maximum continuous load current
12 amperes.
- Choke input filter.
- Line impedance negligible. Peak available fault
current in excess of 1000 amperes.
- Maximum ambient
55°C free convection. Each
semiconductor mounted to a 4" x 4" painted copper
fin ~6" thick.
Requirements for Protective System:
- Protection system must be capable of protecting
=
=
461
SCR MANUAL
CURRENTLIMITING
FUS>:
120 VAt
60 H,
CIRCUIT
BREAKER
1
0
II
'""'"
0
FILTER
CHOKE
C
LOAD
FIGURE 15.14 FAULT PROTECTION CIRCUIT
SCR's and diodes against overloads, DC shorts, and
shorting of individual semiconductors. System can
be shut down when any of these faults occurs.
Solution: - Since the current rating of the IN2156 is higher
than the C35 both at steady-state and under overload, the protection, if properly coordinated with
C35, will be ample for protecting the 1N2156 also.
Using the data for a C35 on a 4"x 4" fin given in
Figure 3.4 for a C34 (the C35 thermal characteristics are identical to the C34) and the load current
rating equation in Figure 3.9(e) which applies for
the continuous square wave of current experienced
in a single-phase circuit with inductive load.
125- 55
POL = 0.0083
(0.0083)
0.0167 X 5.1 + 1 - 0.0167 0.4 - 0.35 + 0.2
= 27 watts
maximum peak heating allowable
per SCR on steady-state basis.
From the specifications for the C35, this level of
heating will be developed by 18 amperes peak or
9 amperes average load current at 180 degree conduction angle with a rectangular current waveshape.
Under inductive load conditions, the maximum
steady-state RMS rating of the complete circuit is
equal to the peak rating of each SCR = 18 amperes.
Assuming that faults and overloads will be superimposed on the steady-state equipment rating of 12
amperes, the semiconductor overload rating can be
calculated from Figure 3.9(f).
P _ TJ
OL -
-
T A - POJ) X Re
Re
(t)
+P
OJ)
For example, for 10 seconds the SCR can dissipate
the following power without its junction exceeding
125°C:
462
PROTECTING THE THYRISTOR AGAINST OVERLOADS AND FAULTS
POL
= 125 -
55 - 8 X 5.1
2.2
+ 8 = 21.3 watts/cell
=
=
Average current rating per cell
13.3 amps (from
specification sheet).
Rated bridge output current
2 X 13.3
26.6
amps RMS.
This point and others calculated by the same means are plotted
on the coordination chart of Figure 15.15. Overload ratings achieved
by this technique limit junction temperature to 125°C.
For non-recurrent types of overload as typified by accidental
short circuits and failure of filter capacitors, the SCR is able to withstand considerably higher overloading as specified in the multi-cycle
surge current ratings. A typical point of this kind can be calculated as
follows. At 0.1 second, a time which is equivalent to 6 cycles on the
surge curve, the peak surge current rating of the C35 is 92 amperes.
The RMS bridge rating is 92 -;- y2 = 65 amperes. This curve blends
into ratings determined from the J2t rating below approximately 50
milliseconds. The I~t rating of the C35 is 75 amps2-sec. At .001 second,
the current rating of the SCR is
y75 amps2-sec.!.00l sec. = 274 amps RMS
Below % cycle, the rating of a single SCR and the rating of the bridge
are identical. Thus, at .001 second, the bridge is rated 274 amps
RMS also.
To afford protection against short circuits of the load and shorted
semiconductors in this type of circuit, a current limiting fuse is
required.
TO .BAMPS eONT
=
.OO.------.T"""
~
~
RESET
g;J
r-
CR8-INI776
CR9-IN1767
CR10-IN5060
~
......
SCR3-G-E CI06Y
04- G-E 2N2646
RI7 - 3300
R18-1 MEG POT,2W.
RI9 - IK, 1/2 W.
R20 - 220, 1/2 W.
R21-0.0070.5W.
FOR OTHER PARTS, SEE FIG. 12:23
C6 - 0.04 MFD
C7 -100 MFD, 30WVDC G-E 62F403
~
c
en
>
Z
C
~
c:
FIGURE 15.18
PHASE CONTROLLED D.C. POWER SUPPLY WITH OVERCURRENT TRIP
!:i
en
SCR MANUAL
current fast enough, a form of electronic crowbar circuit shown in
Figure 8.20 can be very useful. When a fault condition develops in
the load, the crowbar circuit will shunt away the fault current in a
few microseconds for a finite time interval until the interruption of the
fault current can.be performed by conventional means such as by circuit
breaker or fuse.
REFERENCES
1. "Protection of Electronic Power Converters," AlEE Subcommittee
on Electronic Converter Circuits, New York, 1950.
2. "Rectifier Fault Currents," C. C. Herskind, H. L. Kellogg, AlEE
Transactions, March 1945.
3. "Rectifier Fault Currents - II," C. C. Herskind, A. Schmidt, Jr.,
C. E. Rettig, AlEE Transactions, 1949.
4. "Fuse for Semiconductor Protection a Special Breed," M. Goldstein,
1970 IEEE IGA Conference Record, Vol. 70-Cl, October 1970.
5. "Fuse Protection for Power Thyristors," E. T. Schonhlzer, 1970
IEEE IGA Conference Record, Vol. 70-Cl, October 1970.
6. "Application of Fuses for the .Protection of.Diodes and Thyristors,"
K. Lerstrup, 1970 IEEE IGA Conference Record, Vol. 70-Cl,
October 1970.
7. "Fuse Coordination With Power Semiconductors," F. B. Golden,
Paper presented at IEEE International Convention, New York City,
1968.
8. "Take the Guesswork Out of Fuse Selection," F. B. Golden, The
Electronic Engineer, July 1969.
9. "Application of Fuse With Power Semiconductors in Direct Current
Circuit," P. C. Jacobs, Jr., 1970 IEEE IGA Conference Record,
Vol. 70-Cl, October 1970.
VOLTAGE TRANSIENTS IN THYRISTOR CIRCUITS
16
VOLTAGE TRANSIENTS IN THYRISTOR CIRCUITS
In an SCR controlled power system, in order to fully utilize the
SCR's capabilities, it is essential to protect SCR's against effects of
overvoltages whether they may be of a transient or long-time duration
nature. A good working knowledge of the magnitude and energy content of overvoltage in the system may often spell the difference between
success and failure of the application. Any switched energy storage
system is a potential source of overvoltage. If the voltage is high
enough above the blocking voltage of the SCR, destruction may follow
either from energy initially stored in the system or by the fault current
which follows as a consequence of the breakover. A number of common voltage transients occur in electrical systems including lighting
surge, switching transients from elsewhere in the system, switching
transients within the control elements themselves, and regenerative
voltages, to name a few. The effects of overvoltages on an SCR can
be either degrading or catastrophic. A catastrophic failure usually manifests itseH immediately upon the incidence of the overvoltage to the
SCR. However, it is also possible for degradation of the SCR to occur
causing latent defects resulting in failure at some future time. Consequently, for system reliability as well as economic reasons, it is a good
design practice to provide the correct means of preventing possible
overvoltages from damaging the SCR's. This can be accomplished by
operating SCR's well below their voltage ratings to provide a factor of
safety against long time duration overvoltages and by using additional
circuit elements to suppress transient overvoltages at the SCR terminals
to a safe level.
Because of the profound inHuence of voltage transients on successful and reliable operation of SCR circuits an understanding of the
sources of transient voltages and the means of reducing them is essential. Thoughtful design practices can then achieve optimum and
economical use of the ratings of semiconductor components.
16.1 WHERE TO EXPECT VOLTAGE TRANSIENTS1,2,3
In the following discussion transients are considered to be those
voltage levels which exceed the normal repetitive peak voltage applied
to the semiconductor components. In the more common rectifier circuits
operating from an AC source, the repetitive peak reverse voltage
(VROM) applied to the semiconductors is equal to the peak line-to-line
voltage feeding the circuit. In inverter circuits and other types of DC
switches, the repetitive peak voltage applied to SCR's is a function of
the particular circuit and must be analyzed on an individual basis.
Either or both forward and reverse voltage may change widely in
normal circuit operation as load current, conduction angle, load power
factor, etc., are varied.
469
SCR MANUAl
In general, the effect of transient voltages on SCR's and other
thyristors is similar to their effect on conventional silicon rectifier
diodes, but it should be kept in mind that a thyristor is capable of
acting as a high resistance in the forward direction as well as the
reverse. In some instances, this blocking action will prevent transient
energy from being delivered to and dissipated in the load unless the
thyristor first breaks over in the forward direction.
In addition to random line disturbances such as lightning which
have been recorded as high as 5600 volts on a 120-volt residential
power line, transient voltages across thyristor circuits may be generated by occurrences such as those described in Figures 16.1 through
16.8. The indicated power semiconductors may be rectifiers, thyristors
or a combination of both as shown.
OPENING
SWITCH
~-
OJ
...... SWITCH
OPENED
C
WLTAGE
LOAD
FIGURE 16.1
VOLTAGE TRANSIENT DUE TO INTERRUPTION OF TRANSFORMER
MAGNETIZING CURRENT
LINE
VOL~:G~
CLOSING
SWITCH
U~ls
/\
V:H\!
~CLOSED
II
L
SECONDARY
VOLTAGE
Vs
LOAD
FIGURE 16.2 VOLTAGE TRANSIENT DUE TO ENERGIZING TRANSFORMER PRIMARY
470
VOLTAGE TRANSIENTS IN THYRISTOR CIRCUITS
/\
\TV
LINE
VO~:~
FIGURE 16.3 VOLTAGE TRANSIENT DUE TO SWITCHING CIRCUIT WITH INDUCTIVE
LOAD ACROSS INPUT
SOURCE OR
TRANSFClRIlER LEAKAGE
REII;T~7
=
OPENING
SWITCH
+
FIGURE 16.4 VOLTAGE TRANSIENT DUE TO LOAD SWITCHING
CLOSING
~ITCHO""C
INTERWlIIDING
CAPACITANCE
/
--If-I
I
t
Vp
IlL
I
_~~J
FIGURE 16.5 VOLTAGE TRANSIENT DUE TO ENERGIZING STEP-DOWN TRANSFORMERS
471
SCR MANUAL
.Jt!:-f\/\
LEAK7
SOURCE OR
TRANSFORMER
REACTANCE
~
RECOVERY PEAKS
v..
FIGURE 16.6 CYCLICAL COMMUTATION TRANSIENT DUE TO REVERSE RECOVERY OF
SCR's AND RECTIFIER DIODES
CURTH:~NT./'
~KEO
t
LOAD
CURRENT
____L-______-r~
__________
I
VN:
WLTAGE
Ivp
~~~O~
____________~~__~________
Vc
OPENING
SWITCH
UNDER
LOAD
0
VOLTAGE
ACROSS
DC OUTPUT
L-J\J"",..--' TERMINALS OF
RECTIFIER
LOAD
VDC
n------
FIGURE 16.7 VOLTAGE TRANSIENT DUE TO DROPPING LOAD FROM EL·nPE FILTER
WITH HIGH LlC RATIO
FIELO
FIGURE 16.8 OVERVOLTAGE DUE TO REGENERATIVE LOAD
472
VOLTAGE TRANSIENTS IN THYRISTOR CIRCUITS
16.2 HOW TO FIND VOLTAGE TRANSIENTS
Sometimes the presence of excessive voltage transients in an SCR
control circuit is first suspected because of a rash of semiconductor
failures in the prototype equipment in the laboratory. Worse yet, these
first symptoms sometimes wait until the first equipment is shipped
into the field where operating conditions, may on occasion, differ
from the conditions that had been successfully passed in the laboratory. When these failures occur at very light loads or immediately
following circuit switching, voltage transients should be suspected as
the culprit.
Since the search and measurement for possible voltage transients
in a circuit may destroy or at least permanently harm semiconductors
in the circuit, the anode supply voltage should be reduced to about
% or % the normal level initially and then gradually increased as
measurements indicate the absence or reduction of transients to levels
that the semiconductors can withstand.
AC switching transients are usually worst at no load. Therefore,
it may be desirable to test the circuit for this type of transient at no
load with semiconductors of a lower current rating subsituted for the
main devices in order to reduce the cost of components that may be
destroyed in the course of the test. Also, the higher blocking resistances of lower current components will aggravate voltage transients
and thus will generally make measurements and corrective measures
conservative.
16.2.1 Meters
Except for very slow high energy transients, instruments with
moving coils as their detecting and indicating means are almost useless in measuring transient voltages because of their high inertia and
low input impedance. Of the several transient voltage problems discussed earlier, this type of meter may be useful only in measuring the
amplitude of regenerative voltage transients such as those generated
by a hoist motor being driven by an overhauling load.
16.2.2 Oscilloscopes
A high speed oscilloscope with long persistence screen is probably
the most useful single tool for analysis of voltage transients. For significant results in detecting and measuring all the types of transients
that may cause SCR failure, the oscilloscope should have a transient
response of at least 0.1 microsecond rise-time and be capable of
writing rates in excess of ten million inches per_ second. Many commercial oscilloscopes meet this specification. A practical screen material
is the Pll phosphor. Storage or memory scopes, although handicapped
by relatively slow writing speeds and rise-times, are very useful for
recording the longer duration types of transients.
For looking at cyclical transients such as those due to reverse
recovery effects as discussed in Figure 16.6, the use of a scope is
straightforward. In this case, the sweep should be repetitive and synchronized with the power system. However, for nonrecurrent types of
473
SCR MANUAL
transients -due to switching, more careful precautions are necessary.
The scope should be equipped with a hood, and for visual inspection
the room should be -darkened if possible and the eyes of the operator
permitted to become accustomed to a low light level. For checking
the amplitude of voltage transients visually, it is sometimes more effective not to use a horizontal sweep, but to use instead only the vertical
deflection of the trace. Thus, the eyes can be focused on the precise
part of the scope face where the transients will appear, if and when
they occur.
When a sweep is employed, it can be. triggered by the transient
itself or by some external means such as an extra contact or interlock
on the circuit switch which initiates the transient. By this latter means,
the sweep can be initiated before the transient occurs and any doubt
about missing an early part of the transient is eliminated.
The objectiveness of studying and measuring non-cyclical types of
transients is enhanced if a photographic record is secured in addition
to the fleeting image recorded in the mind by the human eye. In many
cases, fast film such as Polaroid Type 42, 44, or 47 (exposure index
200, 400, and 3000, respectively) will catch traces that are not perceptible to the eye.
Circuits should be checked for possible destructive voltage transients by connecting the scope input directly across the semiconductor
to be checked.
16.2.3 Peak Recording Instruments
Electronic peak recording instruments with a memory can be
very useful in checking for transients when their occurrence is random and cannot be predicted.
The ideal voltage measuring equipment for this purpose should
indicate amplitude, waveshape and duration, and frequency of occurrence of overvoltages while recording and maintaining this record over
long periods 'of time while unattended. This ideal combination of characteristics is extremely expensive, and difficult if not impossible, to
secure in a commercially available instrument.
A simple and easy-to-buildinstrument of this type is discussed
here. The primary features of the transient voltage indicator described
here are its:
1. Accuracy and Sensitivity ... 2 % of "full scale" down to 1
".second pulse duration.
2. High input Impedance ... 1 megohm shunted by 5 JL,J.
3. Wide Voltage Range ... dependent on voltage divider design.
4. Unattended Operation ... retains record of transient occurrence up to 12 days.
5. LowCost
6. Battery Operation ... portable, unaffected by line disturbances.
Records transients caused by power failures, and maintains
reading through power failures. Can be operated above ground
potential.
The transient voltage indicator acts as a "go-no-go" type instrument. The user presets a level of voltage on the precision potentiometer dial. If this instantaneous voltage is exceeded, the circuit energizes
474
VOLTAGE TRANSIENTS IN THYRISTOR CIRCUITS
the indicating lamp which remains lit until the circuit is reset by pushing the reset button.
The electrical circuit shown in Figure 16.9 employs a unijunction
transistor, Q1, to compare the input signal with the reference and to
actuate the tripping and latching circuits. The unijunction transistor
CONTACT ON COIL A OF LRI
REGULATED DC
COAXIAL
SIGNAL INPUT
R7
.------{O)
CI
C3
C4
Cs
(ALL CONTACTS SHOWN IN POSITION AFTER RESET)
SCR-GE C220F CONTROLLED RECTIFIER
01 - GE 2N490 UNIJUNCTION TRANSISTOR
02- GE 2N3416 TRANSISTOR
CRI-GE DT230H RECTIFIER
CRZ-Z4XISB ZENER DIODE,17-ZIV
BI- Z2 112 VOLT SATTERT, BURGESS 4156
8;!- I 112 VOLT BATTERT, BURGESS ZFBP
RI-50,OOOQ HELl POT, SERIESC, 3 TURN
RZ-4700Q liZ WATT
R3-470Q liZ WATT
R4-IOOQ 112 WATT
R5-4700Q 112 WATT
R6,R7-47Q I WATT
Rs-zoo£lIWATT
R9-Z500QIWATT, I %TOLERANCE
RIO,R,,-499,OOOQ IWATT,I% TOLERANCE
LRI-POTTER-BRUMFIELD LATCHING RELAY TYPE KEI7D-12VDC
II-GE TYPE 49 LAMP BULB, .06A, Z VOLTS
CI-100 MFD, 50V DC ELECTROLYTIC CAPACITOR GE 76FOZLNIOI
ez-2.oMFD, ZOOV DC PAPER CAPACITOR GE BAI7B205B
C3-Z000 MMFD MICA CAPACITOR
C4, CS-I TO 7.5 MMF CERAMIC TRIMMER CAPACITORS
FIGURE 18.9 CIRCUIT DIAGRAM OF TRANSIENT VOLTAGE INDICATOR
is an ideal device for these functions since it has a very stable Dring
point and presents a high impedance to signals below its tripping
voltage.
The input signal at which the unijunction transistor Q1 trips is
set by potentiometer Rl. Figure 16,10 shows a calibration chart which
~
•
~
170
.10
1000200 IOOIODO 1400 I8DO
400 800 J200 lSOO2Ooo
FIGURE 18.10 CALIBRATION CftART FOR TRANSIENT VOLTAGE INDICATOR
475
SCR MANUAL
defines the input tripping voltage in terms of the potentiometer dial setting. When the unijunction trips, it fires a silicon controlled· rectifier
SCR, thereby actuating a latching relay LRI and lighting an indicating
lamp II in a separate low voltage circuit. At the same time the latching relay de-energizes the tripping circuit from its battery to shut off
the controlled rectifier and conserve battery energy.
Depressing the "reset" button energizes the other coil in the
latching relay, extinguishing the lamp and readying the circuit for
the next trip.
Transistor Q2 in conjunction. with reference diode CR2 applies
a regulated DC voltage to the unijunction transistor and its bias circuit.
Thus battery voltage fluctuation with life do not affect the accuracy
of the circuit until the battery voltage drops below the avalanche voltage of CR2.
The voltage signal to be monitored by the equipment is introduced
at the coaxial signal input and, depending on the desired voltage range
of the instrument, is stepped down by a suitable voltage divider. In
this instrument, the signal introduced to the unijunction circuit across
resistor R9 is 1/400th of the input voltage, having been stepped down
by the RC network consisting of R9, RIO, Rll, C3, C4, and C5.
Capacitor C2 provides sufficient energy for triggering the SCR.
When QI triggers, C2 discharges through diode CRI, QI, and
resistor R4.
Tests on the equipment described here showed a maximum error
of the dial reading using a 3-tum. precision potentiometer that was no
greater than 2 % of "full scale" for pulses from 1 ,...second duration up
to pure DC. Thus, with this 2000 volt instrument maximum error was
40 volts. For pulses of shorter duration than 1 ""second, the error
increases. At 1h ,...second, the maximum error is approximately 5%.
These tests were conducted with a square wave. pulse generator
furnishing the signal. For peaked voltage waveforms, the instrument
reads the voltage level at which the waveform is approximately
1h ,...second wide.
While this instrument has accuracy well beyond instruments
many times its cost, and is ample for the purpose intended, it is felt
that a substantially higher level of accuracy could be incorporated by
using more precise components, a more stable voltage supply, and a
better optimized voltage divider.
16.2.4 Spark Gaps
For high voltage systems, calibrated sphere spark gaps can be
used to measure the crest values of transient voltages. Current through
the spark gap after it has broken down should be limited by a noninductive resistance (at least one ohm per volt of test voltage) in series
with the gap on the grounded side. Suitable overcurrent protective
devices should be used to interrupt the power follow-through after the
voltage surge has passed. In general, the breakdown voltage level for
gaps varies significantly with the waveshapeofthe voltage being meas~
ured as well as with many environmental factors.
476
VOLTAGE TRANSIENTS IN THYRISTOR CIRCUITS
16.3 SUPPRESSION TECHNIQUES
Three basic approaches can be employed for the suppression of
transient overvoltages:
1) Series suppression
2) Shunt suppression
3) Combination of 1 and 2
Functionally, series transient suppressors act as .a series impedance which varies from a low resistance under normal operating
conditions to a high resistance when a transient appears; shunt transient suppressors appear as an open circuit under normal operating
conditions and become a low-impedance shunt path during a transient.
Two distinct advantages of using shunt suppressors in comparison with
series suppressors are the absence of insertion loss and simplicity of
the circuit arrangement. However, since it presents a low impedance
to the transient, the shunt suppressor must be able to absorb large
amounts of power for a short duration of time under repetitive pulsed
conditions.
The most practical method of suppressing voltage transients is
the third approach. By utilizing the available circuit inductance
together with a properly designed shunt suppressor, a series-shunt
suppression network is formed with the combined advantages of series
and shunt suppression. Such a technique will be discussed later.
In general, voltage suppressors may be grouped into two distinct
categories: suppression components and suppression networks.
16.3.1 Suppression Components4,11
There are many types of voltage suppression components available on the market today. This section will discuss two of the more
popular ones applied in SCR controlled power circuits.
16.3.1.1 Polycrystalline Suppressors:
Selenium Thyrectors and Metal Oxide VaristorS
Comparable samples of two families of polycrystalIine suppressors
available to the designer are shown in Figure 16.11. Selenium Thyrectors evolved from selenium diode rectifier technology in the late fifties
and early sixties. When arranged in a bipolar configuration by stacking
plates of opposite polarity together in a series electrical arrangement
the V-I characteristic is as shown in Figure 16.12. While selenium
Thyrectors consist of an integral number of plates each having a fixed
maximum operating voltage at steady-state conditions as shown on
Point A of the curve, General Electric GE-MOVTM Metal Oxide Varistors are fabricated from a ceramic powder by a pressing operation.
GE-MOV varistor characteristics depend upon bulk action within the
ceramic of. the crystal structure. Urilike the selenium suppressor an
effective continuous variation of characteristic voltage rating can be
achieved by pressing to controlled dimensions (thickness, length, etc.).
Similarly to selenium, average power handling capability as well as
pulse energy capability is determined by the diameter of the GE-MOV
varistor and means used to cool it.
477
SCR MANUAL
Referring again to Figure 16.12, the slope of the V-I characteristic beyond the maximum rated voltage Point A defines the clamping
performance of the suppressor. This slope together with the so-called
knee of the linear characteristic is best defined by redrawing the V-I
curve on log-log scales as shown in Figure 16.13 where the characteristics of three different families of suppressors plus a resistor are given.
The curves can all be expressed as:
1= KVex
where
K is a device constant
ex is the slope of the curve on log-log scales and is defined as:
ex=
where
Log (12 /11 )
Log (V2 /Vt )
11 and 12 are taken a decade apart
,
$
THYRECTOR
GE-MOJ3
FIGURE 16.11
TWO FAMILIES OF POLYCRYSTALLINE SUPPRESSORS
1+)
FIGURE 16.12
478
NON·POLARIZED VOLT-CURRENT CHARACTERISTIC OF PDLYCRYSTALLINE
SUPPRESSORS
VOLTAGE TRANSIENTS IN THYRISTOR CIRCUITS
,
RESISTOR
(oe!! Il
en
1000
o
800
~
~
/
1&1
/
500
g
400
'"o
300
J
" .
~.--
1&1
Z
j:!
/
6" THYRITE VARISTOR
7J:f~~--r-
,/
(!)
~
7V
200
~
v
- ---
--- - .r
---rr-T'n'
'T
iT
1
-
1"I'SEL~oei~
l-
V r > 2 5 1 - f-----
m
~
OU00
100
I
2
3
4
5
8
10
20
30
40 50
80100
INSTANTANEOUS CURRENT (AMPS) -
FIGURE 16.13
LOG·LOG VOLT/AMPERE CHARACTERISTICS OF COMMON VOLTAGE SUPPRESSORS
Both selenium Thyrectors and GE-MOV varistors are thus seen to
be voltage dependent symmetrical resistors having a high degree of
non-linearity. Terminal impedance at voltages lower than nominal are
very high while impedance progresses to an extremely low value as
voltage is increased to a model's upper range, providing the desired
suppressing or clamping function. Thus polycrystalline suppressors
provide a means of absorbing transient energy pulses while limiting
the rise of transient voltage in the circuit to controlled levels.
An examination of a typical SCR controlled circuit will illustrate
the usefulness of such a device. Consider the single phase bridge circuit in Figures 16.1 and 16.2. Transient voltages may result from
either switching on or off the primary of the transformer. The voltage
applied to the SCR AC terminals is the voltage appearing at the transformer secondary winding. If the transformer magnetizing current is
interrupted at its peak, the SCR's may be subjected to a transient voltage up to ten' times the normal peak secondary voltage of the transformer (Figure 16.1). On closing the primary circuit at the peak of the
voltage wave SCR's may be subjected to a voltage transient double
that of the normal peak secondary voltage of the transformer
(Figure 16.2).
By connecting a suppressor across the secondary winding of the
transformer, the voltage transient appearing at the SCR AC terminals
will be limited. In order to assist engineers in selecting the proper
suppressor for this application, the following design outline is given:
Suppressor Selection Guide for Transformer Circuit Applications:
1. What voltage should be used?
a) Determine maximum steady state RMS voltage that will be
applied to the suppressor and specify the same or closest
479
SCR MANUAL
higher rating. For non-sinusoidal voltages, use recurrent
peak voltage ratings.
2. What clamping voltage will be obtained?
a) Determine maximum peak current that will be carried by
the suppressor. In transformer circuits this is the peak magnetizing current (iM) X transformer turns ratio. For most
cases assume iM =. IE X y2 where IE = exciting current
which is the no load RMS input current read at max. RMS
design voltage. Check to see that iM does not exceed rated
peak current value of suppressor chosen.
b) Using Figure 16.14, read the suppressed peak voltage ratio
at peak current and multiply times the recurrent peak voltage spec of the chosen model in. the max. rating table of
the pertinent data sheet.
3. What energy rating is required?
a) E = lh ~iM2 where LM is the equivalent magnetizing
inductance.
b)
T
-
"-'M -
XLM h
X - Primary Voltage (RMS) d
2,rf were LM I
an
.
]I[
ill[ )
( 1]1[= y2
c) Choose model with an energy rating higher than the calculated value.
4. What is the power dissipation?
a) For repetitive pulses:
Avg. Pwr Energy/pulse X Rep. Rate
b) Make sure the specified power capability of the device is
not exceeded and is properly derated.
=
2.4
0
;:::
2.2 I--
~f2E5~!~~~QUARE) L
TRANSIENT CONDITIONS
I--
PEAK AMPERES VS
SUPPRESSED PEAK VOLTAGE RATI
~~-MO~ V~Rllsi6~
.i
o
(19.:10 &26.21 M M
DIAMETER)
/""',
'"
IE
"'"~
j
2.0
0
«"
Q.
Q
~
1.8
'"
'"
Ul
Ul
IE
Q.
Q.
~
Ul
~
1.6
--L
V
RECURRENT
PEAK VOLTAGE
SUPPRESSED
PEAK PULSE VOLTAGE
PEAK VOLTAGE" RECURRENT PEAK
RATIO
VOLTAGE RATING
I
I
I I 1111
~UM
A,
~
-'
1.5
1.4
f-'
~
.1
III
/
I
>
'"
I I I III
1.0
10
I I 111111
1
100
PEAK INSTANTANEOUS CURRENT (AMPERES)
FIGURE 16.14 POLYCRYSTAUINE SUPPRESSION RATIO CURVES
480
I I 11111
1000
VOLTAGE TRANSIENTS IN THYRISTOR CIRCUITS
Table 16.1 lists the major design parameter capabilities of two
comparably sized polycrystalline suppressors. Larger sizes are available having higher transient energy handling properties. For further
information consult device data sheets.
Non-Repetitive
Max Average
Energy Dissipation Power Dissipation
Watt-Seconds
Watts
1;110
RMS
Voltage
Rating
130
250
480
II
I; II
II
,> "0
• >~~ • 11~~ ~O""o
• ti~o ~~
...... ~It::;:: ~'"
""j:ap..~~~~~~
30
30
30
1000
1000
1000
%,"
1"
Thyrector MOV
12
12
12
1
1
1
COMPARISON OF 1" THYRECTORS WITH 0/4" IE·MOV VARISTORS
Being able to limit voltage transients to a controlled value, a
circuit designer can add value to the application by increasing reliability .and life of the components protected thus providing additional
margins of safety and reliability to the equipment and user.
Although suppression components, such as Thyrector diodes, metal
oxide varistors and others, are effective in clipping voltage transients
to a designed level, they do very little to limit the rate of change of
the transient voltage. Since SCR's are sensitive to the forward applied
dv/dt, if an excessive rate is applied, it may turn-on without having a
gate signal applied. The spurious turn-on may even occur though the
applied forward voltage amplitude is considerably below the rated
peak anode voltage VDRM of the SCR. Such an unscheduled turn-on
may result in excessive surge current which can cause SCR's to fail.
For dv/dt protection, consequently, normal suppression components,
such as Thyrector diodes, metal oxide varistors, controlled avalanche
diodes, or spark gaps will not be sufficient. A properly design resistorcapacitor network will not only limit the dv I dt to a desired level, it can
also aid in reducing the repetitive peak transient voltage to a more
practical value.
16.3.2 Suppression Network
One form of suppression network is commonly called a snubber
circuit. The snubber circuit basically consists of a series-connected
resistor and capacitor placed in shunt with an SCR. The snubber
circuit in conjunction with the circuit effective series inductance controls the maximum rate of change of voltage and the peak voltage
across the device when a stepped forward voltage is applied to it.
Referring to Figure 16.15 when an input is suddenly applied, it is transiently divided between the inductance, L, which functions as a series
suppressor and the R-C snubber circuit.
481
SCR MANUAL
16.3.2.1 Snubber Calculation for D.C.Circuit6
When a circuit designer works with power SCR's in designing a
snubber, he is likely to use a cut-and-try method. Such a technique can
be tedious and time consuming. By using designed nomograph, Figures
16.16 through 16.18, the various trial steps can be eliminated. The
construction of these nomographs is based on the analysis of a basic
R-C snubber circuit in response to a step input signal. The analysis
shows that the effect of damping in a L-R-C circuit can be described
in terms of a single parameter; designated £, which is the ratio of the
resistance to the surge impedance of the circuit. The effective total
circuit inductance is normalized in terms of £, R, and C of the circuit.
The relationship is shown by the following expression:
£=
-it2 yC/L
(16.3)
It is desirable to have a high value of L. A higher value of L will
allow higher value of R and a lower value of C to retain the desired
damping effect, controlled dvldt and peak overshoot voltage (these
relationships are clearly expressed by Figures 16.16 and 16.17. Yet a
higher value of R and a lower value of C not only minimizes the power
dissipation in the snubber circuit, but also limits the initial current
discharging into the SCR during. its turn-on interval. Based on experience and test, it is desirable to select a circuit damping ratio, £, in the
range of .5 to 1.0, both to limit the peak overshoot voltage applied to
the SCR and to minimize the "ringing" of the L-R-C circuit within the
maximum required dv I dt value.
SCR
FIGURE 16.15 EQUIVALENT CIRCUIT OF SNUBBER
1.0
2.0
0.8
,,-.
R~
E.
0.6
~
F""---~------,,!i"'"
I.'
0.4
1.4
1.22
0.2
I ••
° _'----_-'-_---'__ 0:::.8:::.5---'-_ _
°
0.2
0.4
0.6
0.8
1.0
f ./CiL.
..LJ.:
.L...-_..........
•••
DAMPING FACTOR. { .
FIGURE 16.16 NORMALIZED PEAK SNUBBER CURRENT AND OVERSHOOT VOLTAGE Vs
CIRCUIT DAMPING RATIO
482
VOLTAGE TRANSIENTS IN THYRISTOR CIRCUITS
DAMPING RATIO
~
1.0
0.'
0 ••
0.7
0.•
0.5
0.4
Icr'
0.'
'0·
0.2
10'
0.1
10·
[~ • 4~·RC'~·
FOR
oS~ < I
[~. 4~J'RC"
FOR~"'O
FIGURE 16.17 NORMALIZED SNUBBER CIRCUIT TIME CONSTANT SELECTION CHART
c.uId
2.0
Hz
'0 '
E.
2000
1.0
1000
®
0.1
100
0.01
0.005
1,,1I1! I
10'
I
L"" ~ I
10
I
102@
10.64 2
WATTS
I
FIGURE 16.18 SNUBBER LOSS NOMOGRAPH
Worked Example
=
Assume: 1. Peak switching voltage, Es
600 volts.
2. Operating frequency, fo 400 Hz.
3. dv/dtlm to he limited ~ 500 V I "sec.
4. Chosen ~ = .65 for a controlled voltage overshoot of
approximately 22% (Figure 16.16).
Solution: 1. To determine the required R-C time constant of the snubber, go to Figure 16.17, connecting two points specified
=
by:
P' A. dv/dtl m _ 500 """ 83
OInt . Es - 600 - .
Point B: ~ = .65
R-C is located (Point C) to he 2.0 p.Sec.
2. To determine the value of R, go to Figure 16.16 and
locate: R
Ip Es
.63
=
483
SCR MANUAL
Assume the
to be: R
IP(Snubber)
is limited to 50 amperes,R is found
= (.63) (~~) = 7.6 O. Use 8 o.
3. From Steps 1 and 2, C is determined
C
= ::~ = .25 pFd.
4. Based on C, Hz and E s , go to Figure 16.16 and find the
peak snubber power dissipation. By connecting two points
specified by:
Point D: C = .25 pFd
Point E: Es
600 volts
Point F (Je = .045 Joules) is located.
Project Point F horizontally to "Hz" scale on Point G,
then vertically project G to watts scale on Point H. The
maximum power dissipation is found to be 40 watts.
=
It has been stated that SCR's can be affected by voltage transients from the AC power supply. Voltage transients from the power
supply are primarily caused by the effects of switching inductive circuits such as are always present in the supply transformer. Consider
the single-phase bridge circuit in Figure 16.19. Transients may result
from switching off the primary of the transformer. If a non-inductive
load is always connected and the load is able to absorb sufficient
energy to attenuate the induced voltage, no transient suppression
measures are required. However, if appreciable inductance is present
in the load and no-load operation is possible, transient suppression is
a must. A properly chosen R-C snubber connected across the secondary of the transform will dampen the transient voltage to a desired
level. The size of resistor-capacitor required for a particular suppression job is a function of many circuit parameters such as the type of
load, load current level, the transformer characteristics and the frequency of interruption and switching.
16.3.2.2 Snubber Calculation for AC Circuit6,7,8
The following equation has proved useful in selecting the required
capacitance sufficiently to limit the voltage transients within SCR voltage ratings:
VA 60 .
C
10 VS 2 ""[mIcrofarads
(16.4)
=
where
C = the minimum required capacitance
VA
the transformer volt-ampere rating
Vs
the transformer secondary RMS voltage
f
frequency other than 60 Hz
=
=
=
The required resistance to ensure adequate damping can be calculated
from the following relationship:
R
2 (y'L/C
=
484
VOLTAGE TRANSIENTS IN THYRISTOR CIRCUITS
where
R = the required resistance to damp the transient voltage to
a desired level
£
damping factor
L = effective circuit inductance
C minimum required capacitance
=
=
WITHOUT SUPI'RUSOR
WITH IUPPIIEUOR
t
(a) Energizing a Transformer Primary
I"
v......._
v.
I
(b) De-Energizing a Transformer Primary
FIGURE 16.19 VOLTAGE TRANSIENTS DUE TO SWITCHING Of TIIAIISfORMER PRIMARY
Work Example - Refer to Figure 16.19
Assume: 1. Supply transformer is rated 5 KVA with secondary voltage of 120 VRMS
2. Switching frequency, fo
400 Hz
3. Circuit inductance, L
100 phy at 400 Hz
4. Peak transient voltage is to be limited at 200 volts
Solution: 1. To determine required snubber capacitance, using
formula
=
=
C = 10 VA 60
Vs f
5000 60
= 10 (120)2 400
= .52 pfd -
use .5 pfd
2. Calculate peak transient voltage vs peak switching voltage ratio
Vp
V Sl)
200
120 y2
= 1.18
Go to Figure 16.22 and locate £ = .75
485
SCR MANUAL
3. To determine required damping resistance, using following relationship
R
= 2 (£) y'LiC
= 2 (.75) ~.1.~0
= 21.2 0 - use 20 0
4. To determine the maximum power dissipa,tion, go to
Figure 16.16 using same procedure as outlined in Step 4
of the previous sample with following specified parameters
C = .5 pId
Es = y'2 (120) = 170 volts
fo = 400 Hz
The maximum power dissipation in the resistor is found
to be 10 watts.
To suppress the voltage transients generated from the power supply of a single-phase system, the snubber should be connected across
the secondary of the transformer.
For a three phase system, the snubber should be connected either
from line to line or line to neutral across the secondary of the transformer depending upon the secondary connections whether it may be
of delta or star configurations.
16.4 MISCELLANEOUS METHODS
Several other transient suppression means may be used to good
advantage depending on the particular circumstances of the application. Spark gaps may be used in high voltage circuits provided the
precautions outlined in Section 16.2.4 are maintained. 6 Silicon diodes
can be used as discharge paths for the energy stored in inductive circuit elements such as generator fields and magnetic brakes.
Electronic crowbar circuits of the type shown in Figure 8.20 use
the SCR to provide microsecond protection against overvoltage conditions for entire circuits. Properly selected and applied triac and diac
components can also be used to shunt transient energy away from
sensitive electronic circuitry when voltage tries to rise above the breakover switching level of the particular protective semiconductor
component.
Figure 16.20 illustrates a technique by using SCR's to control
dynamic braking to a DC motor load, thus preventing the· occurrence
of overvoltage from damaging the motor and subsequently limiting
the DC voltage imposed on SCR's in the power circuit. Under normal
operation neither SCR1 or SCR2 conduct and no energy will be dissipated in Rl and R2. When the motor CEMF voltage rises above a
level predetermined by the selection of avalanche diodes CRl and CR 2,
SCR 1 and SCR 2 are triggered, connecting Rl and R2 across the load.
As soon as the motor CEMF voltage drops below the controlled supplied voltage, SCR 1 and SCR2 will be commutated off by the AC line
486
VOLTAGE TRANSIENTS IN THYRISTOR CIRCUITS
POWER UNIT
V A.C.
--++--..
M
----LOAD
FIGURE 16.20 REGENERATIVE VOLTAGE PROTECTION
and return to their non-conducting state. The circuit performs a dual
function of limiting overvoltage and overcurrent in a continuous mode
of action. Such a function provides an attractive feature in motor control applications where non-interruptive performance is demanded.
In general, if the protective network is properly designed, the
system reliability can be greatly improved. In order to design the
protective network properly, a designer must know the source and
nature of possible transients in his circuit and the characteristics of
the control elements to be protected.
REFERENCES
1. "Rectifier Voltage Transients - Causes, Detection, Reduction,"
F. W. Gutzwiller, Electrical Manufacturing, December 1959.
2. "Surge Voltage in Residential and Industrial Power Circuits," F. D.
Martzloff and G. J. Hahn, IEEE Transactions on Power Apparatus
& Systems, Vol. PAS 89, No.6, July/August 1970.
3. IEEE Committee Report "Bibliography on Surge Voltages in AC
Power Circuits Rated 600 Volts and Less," IEEE Transactions on
Power Apparatus and Systems, Vol. PAS 89, No.6, July/August
1970.
4. "General Electric Selenium Thyrector Diodes," Application Note
200.5,* General Electric Company, Syracuse, N. Y.
5. "An Introduction to the Controlled Avalanche Silicon Rectifier,"
Application Note 200.27,* General Electric Company, Syracuse,
N.Y.
6. "Analysis and Design of Optimized Snubber Circuits for dv/dt
Protection in Power Thyristor Applications," Publication 660.24,*
General Electric Company, Syracuse, N. Y.
7. "Commutation dv/dt Effects in Thyristor Three-Phase Bridge
Converters," J. B. Rice and L. E. Nickels, IEEE IGA Transactions,
Vol. IGA-4, No.6, November/December, 1968.
487
SCR MANUAL
8. "Practical Transient Suppression Circuits for Thyristor Power
Control Systems," J. Merret, Mullard Technical Communications,
No. 104, March 1970.
9. "Optimum Snubbers for Power Semiconductors," W. McMurray,
IEEE, IGA 1971 Conference Record.
10. "Design of Snubber Circuits for Thyristor Converters," J. B. Rice,
IEEE IGA Conference Record, 1969, pp. 483.
11. GE-MOVTM Varistors, Voltage Transient Suppressors by F. B.
Golden and R. W. Fox, Application Note 200.60,* General Electric Company, Auburn, N. Y.
*See Chapter 23 for availability and ordering information.
488
RADIO FREQUENCY INTERFERENCE AND INTERACTION OF THYRISTORS
17
RADIO FREQUENCY INTERFERENCE
AND INTERACTION OF THYRISTORS
17.1 INTRODUCTION
Each time a thyristor is triggered in a resistive. circuit, the load
current goes from zero to the load limited current value in less than a
few microseconds. A frequency analysis of such a step function of current would show an infinite spectrum of energy, with an amplitude
inversely proportional to frequency. With full wave phase control in a
60 Hz circuit, there is a pulse of this noise 120 times a second. In
applications where phase control is used in the home, such as lamp
dimming, this can be extremely annoying, for while the frequencies
generated would not generally bother television or FM radio reception,
the broadcast band of AM radio would suffer severe interference. In an
industrial environment, where several control circuits may be used,
these noise pulses cause interaction between one thyristor control
and another. The power system can act as a large transmission line and
antenna system, propagating these radio frequency disturbances for a
considerable distance.
With the newer inverter types. of SCR's and their growing use,
an additional problem is created. since the· basic inverter frequencies
may be in excess of 10 kHz. The harmonics.generated,in these systems
can cause interference sources which are orders of magnitudes larger
than those of phase control systems and also the fundamental frequency
over 10 kHz may put the equipment under specific provisions of the
FCC rules.
At this time the Federal Communications Commission rules (Part
15) require that incidental radiation devices " ... shall not cause any
harmful interference in use." It is under this section of the law which
most thyristor systems fall. The Commission has stated· a willingness
to allow industry to police itself. To this end several of the national
associations have proposed standards for their members, It is strongly
urged that anyone desiring to sell or lease, offer for sale or lease, import,
ship or distribute such equipment, ascertain whether or not he is in
compliance with the applicable standards. Included at the end of this
chapter is a list of some of the currently available standards for both
the United States and European countries.
17.2 THE NATURE OF RADIO FREQUENCY INTERFERENCE (RFI)
There are two basic forms of RFI to consider. The :first (and most
commonly measured) is conducted RFI. In this form, the high frequency energy generated by the thyristor switching transients propagates through the power lines, which act as transmission lines. By using
standard methods and equipment, quantitative measurements may be
489
SCRMANUAL
fairly easily obtained on conducted RFI. These standards are listed. in
Section 17.7.
The other main form is that of radiated RFI. This is the RF
energy which is radiated directly from the equipment. This is a difficult
type of RFI to measure since it can never be separated from the problems of location, wiring layout, ground effects, etc.
In most cases, the radiated RFI from a properly designed piece
of equipment is insignificant compared to the re-radiation of conducted
RFI from the large antenna system we call power lines.
The following military specifications set quantitative interference
levels and give test procedures to which an equipment must be qualified if it is to conform to the specification:
MIL-STD-461A (Requirements for Equipment)
MIL-STD-462 (Measurement Methods)
MIL-STD-463 (Definitions)
17.2.1 Filter Design
Since thyristors generate essentially a step function of current
when they turn on into a resistive load, the conducted RFI has the
frequency distribution of a step function, that is, a continuous spectrum of noise with an amplitude which decreases with frequency at a
rate of 20 db per decade. This indicates that even unfiltered thyristor
circuits would show very little tendency to interfere with such VHF
services as television or FM broadcasting. The AM broadcast band
however lies between 550 and 1600 kHz, and would receive severe
interference, if the thyristor circuits were not properly filtered.
The simplest type of filter is merely an inductor in series with the
load resistance to slow the rate of rise of current. This would give a
filter effectiveness of about 20 db/decade. The typical example shown
in Figure 17.1 shows that the bottom of the broadcast band requires
from 40 to 50 db of suppression to reach a level of interference which
could be considered adequate (about 500 quasi-peak p.volts).* To
achieve this, the breakpoint frequency, fo = R/{2n- L) {where R is the
"~
lOOK
r-
If
:r
10,000
~
1000
~
0
~
§
FIGURE 17.1
100
THEORETICAL THYRISTOR CIRCUIT NOISE SPECTRUM WITH AND
WITHOUT FILTERING
·"Quasi-peak Volts" is a unit of measure which is determined by the· standard test methods.
It is in effect a measure of the "Nuisance Value" conducted RFI.
490
RADIO FREQUENCY INTERFERENCE AND INTERACTION OF THYRISTORS
load resistance), would have to be at 5 kHz, or below. This would be
a rather large and costly inductor.
r--'
rh
CL,*
I
I
11---4--'~-... -
L
- -,
T2
=r CH
rh
AC
LINE
AC
LINE
1
L
CL 2
:f
rh
(a) Single Inductor
(b) L·C Filter
C LI ; C L2 = Line Capacitance to Ground
C H = Heatsink Capacitance to Ground
FIGURE 17.2
,--
SIMPLE FILTERS
---------,
,
I
I
I
I
---v C3
"'1'
I
-J.., C4
LI
,
/
_.(
\i
r , m ~~~~~~~~ m
SCRI
SCR2
-r-
I
-T-
AC
I
LINE I
I
I
L2
LOAD
I
I
I
I
.J., C5
CI
/'"1".
L ______
FIGURE 17.3
_
______ J
RFI FILTERING AND SHIELDING FOR BACK·TO·BACK SCH'S
The addition of a shunt capacitance to the filter as shown in
Figure 17 .2(b) gives a far superior characteristic as can be seen in Figure 17.1. Now the required 40 db of suppression can be obtained in
a single decade. As a rule of thumb, the proper values for Land C
may be found by making the L-R and L-C breakpoint frequencies
equal.
1
21T yLC
or in other words
1
27/' fo L
= 27/' fo C = RL
This allows a value of L one tenth that needed for a purely inductive
filter.
491
SCR MANUAL
In a practical thyristor circuit, one side of the device is usually
connected to a heatsink, which because of its size or mounting, is
capacitively connected to ground. In the case of the triac shown in
Figure 17.2 main terminal T 2 is the heatsink side. If the choke L were
in series with T 2 , as in Figure 17.2(a), the heatsink capacitance in
conjunction with stray line capacitance would shunt the choke, thereby
reducing its effectiveness as a filter. The proper connection of L in
series with Tl actually puts the stray capacitances in parallel with C,
thus enhancing filter effectiveness.
The optimum connection for back-to-back SCR's is shown in Figure 17.3. If a shielded enclosure is not present, C 1 and C 2 should be a
single capacitor connected between the anodes of SCR 1 and SCR2 .
It is important to note that any pulse transformers or triggering
circuits should put the smallest possible capacitive loading on the
cathode of the SCR's, since this capacitance will appear across the
chokes.
If you look at the circuits of Figure 17.2, you can see that the
L-C's and triac or SCR pair form a resonant discharge circuit, which
depends on the load impedance for damping. For circuit Q's greater
than about 2.5 the current through the thyristor will reverse, as shown
in Figure 17.4, and a specific triac might tum off if it is a relatively
fast device.
This condition is aggravated for light loads, in this case about
100 watts or less, or somewhat inductive loads, which contribute little
damping to the circuit. The simple L-C circuit does behave properly
however with heavier resistive loads, as shown in Figure 17.4(b). To
obtain proper operation under light load conditions, for instance a
lamp dimmer with a 60 watt lamp, it is necessary to build the damping required into the filter. This can be done by adding another resistor
and capacitor as shown in the circuit of Figure 17.5. The component
values are chosen to give about the same filtering effect as the L-C
filter of Figure 17 .2(b).
,
~
(a) 60 Watt Load
High Q
2.5
>
(b) 150 Watt Load
Low Q
2.5
VERT. - 2 AMP/CM
HORIZ. - 5 pSEC/CM
L = 100 pH
C = 0.1 pF
Supply Voltage
<
= 120 V
FIGURE 17.4 TRIAC CURRENT FOR THE CIRCUIT OF FIGURE 17.2(B) IMMEDIATELY
AFTER THYRISTOR TURN·ON
492
RADIO FREQUENCY INTERFERENCE AND INTERACTION OF THYRISTORS
82
T2
TRIAC
TI
fiGURE 17.5 TYPICAL DAMPED R·F FILTER
17.2.2 Components for R-F Filters
The above discussion centers on the simplest forms of filters. It is
reasonable to assume that more complex filters can do more adequately
if they are required. Figure 17.6 shows the conducted RFI measurements for triac fan motor speed control circuit. The test measurement
was made in accordance with the American National Standards Association's Standard Method C63.4.** Also included on the graph is the
NEMA Standard per WD.2-1970** which sets a maximum allowable
limit for conducted interference.
12 0
.....'"
I I
OJ
:>
60
...
cl' 40
>
oJ
ffi
~
--/
"
NEMA LIMIT
.}
...
...~
...
......'"
D.2V
TRIAC - FAN MOTOR
SPEED CONTROL
(USING ANSI-C63.4 METHOD
OF MEASUREMENT)
100
,. 80
~
I.OV
O.SV
I I II
20
~
- '----
--
-..,
-
0.1 V
!50mY
~
INDUCTOR
REMOVED
FROM CIRCUIT
(CONLY)
\
'\
.......
~
,./
.........,
~L~~i~s~II~~~
0
~
~o:: ~
~
::..-
"' -
5mv
2mV
...~
't-
ImV
iii
sooILv ~
200ILV .J
100ILV ~
SOILV ~
\ ...
-20
10ILV
SILV
IILV
...u
z
......'"
...'"
OJ
~
~
O.IILV
0.15
10
30
FREOUENCY, MHz
FIGURE 17.6 TYPICAL CONDUCTED Rfl WITH AND WITHOUT SUPPRESSION NETWORK
"Refer to end of chapter for ordering information.
493
SCR MANUAL
When choosing components for R-F filters it is extremely important that the components chosen act in the manner in which they
should. At high frequencies capacitors tend to act like inductors and
inductors like capacitors. When this occurs all filtering is lost.
~
"---,
I
I
ZL2
-
rc
:
L ___
s
Z = LlCs
.L2 j(..L-lIwCsl
PRACTICAL
(a) Inductors
LL = 2Ll= ;t~1f'J-Tkij~~
ZCZ=J(wL,t-JlwCI
IDEAL
PRACTICAL
bl CAPACITORS
FIGURE 17.7
FILTER ELEMENTS-IDEAL, PRACTICAL AND THEIR FREQUENCY RESPONSE
Figure 17.8(a) shows how 0.05 pF capacitor's characteristics
change with frequency. Figure 17.8(b) shows the characteristics of a
157 p.H inductor.
0
",~
157,.H
.05Jo1f CAPACITOR
IMPEDANCE
CHARACTERISTIC
-
6" L!ADS/ /
I'~, II
,
K
/
I
I\~\
1\
-
3" LEADS
IRON CORE - TORRoro
3/4"0.0 1/2"1.0
45 TURNS NO. 16 WIRE
NO LOAD IMPEOANCE
VS FREQUENCY
-I
I
1
'\
(MHz
FREQUENCY, I
FIGURE 17.8(A) TYPICAL IMPEDANCE
CHARACTERISTICS OF .05,uF CAPACITOR
_\ ,,
/ /~
k:oRmcAc 1\
I-Ilur,.H
"-
THEORETICAL
o.I
100KHz
1\
V
,
V
V
1/
,
1/
K-
'0
FREQUENCY, MHz
FIGURE 17.8(8) TYPICAL IMPEDANCE
CHARACTERISTICS OF 157,uHINDUCTOR
Besides an inductor maintaining'its inductance at high frequencies,
it should also be designed to prevent saturation from occurring too
soon. When the core of an inductor becomes saturated, the inductor
494
RADIO FREQUENCY INTERFERENCE' AND INTERACTION OF THYRISTORS
acts like it has an air core and its reactance drops thus losing its ability
to lower the di/ dt of the circuit. Most non air core inductors can be
rated for a minimum volt-second withstand capability to guarantee
that saturation does not occur too soon. The volt-second capability of
the reactor should be large enough to provide a load current rise time
of not less than approximately 50 p.sec.
17.2.3 Fast Recovery Rectifiers
In circuits which use rectifier diodes, RF noise may be generated
by the reverse recovery performance of the diodes. Due to the minority
carrier storage effect, the diode does not immediately block voltage
when the circuit causes current reversal. When, after a few p,seconds,
the charge which had been stored in the diode is "swept-out," the
diode can again block the How of reverse current. At this point, the current can stop quite suddenly, giving a "snap-off" effect. The energy
stored in the circuit inductance at this point can be shown to be
equal to
WT
QREo
=
where Q R is the total charge swept-out
Eo is the, circuit commutation voltage.
Figure 17.9 shows the waveform of the recovery current of a conventional as well as a fast recovery diode.
FAST RECOVERY DIODE
....
c
o
5
-I
«
z
o
i=
z
....
>
z
8.__
FIGURE 17.9
COMPARISON OF REVERSE RECOVERY PERFORMANCE OF TYPICAL RECTIFIER DIODES. VERTICAL = 8 AMPS. PER CM.
HORIZONTAL = .5 "SEC. PER CM.
On each "snap-off" commutation of a diode there is a step of
current of height I RP . The RF components of this current step. are
given by
I(w)
I RP
=
7rW
where w is an RF noise component frequency.
This result is found by Fourier Integral Analysis of a step function.
But IRp2 is proportional to W T or
I RP
ex:
yWT
495
SCR MANUAL
Thus, the RF interference generation at a given frequency is
I(w)
yWT
ex:
7TW
Since the value ~f W T runs better than 100 times less for a fast recovery
rectifier than for a conventional rectifier, there is a considerable reduction in RFI problems when using fast recovery devices.
17.2.4 Reduction of Radiated RFI
The minimization of radiated RFI is as much a matter of good
construction practice as anything else. Referring to Figure 17.2(b), the
current through the loop formed by C, L, and the thyristor contains
high frequency components of a much greater magnitude than the line
current. (The inner loop has only a single L filter); The wiring of this
loop can act as an antenna for direct radiation. Since the radiation
efficiency of an antenna of this type is proportional to the area enclosed
by the loop, good practice requires that this current loop be constructed with a minimum of enclosed area. It should be pointed out
that trigger circuits can also be offenders in direct radiation, and the
same techniques apply.
Figure 17.3 illustrates proper shielding techniques. The SCR's
with their filter circuitry are enclosed inside their own shielded compartment, with leads from the power line and to the load passing
through feedthrough capacitors C a-C 6 • * Either the pulse transformer
or the gate pulse generation circuitry should be located within the
compartment, since locating them remotely forces one to hang leads
on the cathodes of the SCR's, thereby providing excellent antennas.
'For devices where "leakage" current criteria must be met the magnitude of these capacitors
may be severely limited.
1000
=
-
BX = sse-nIX..
-
BS - GAUSES AT SURFACE
-
8X- GAUSES AT D~fIITH
'X.- THICkNESS IN eM
-
t:*p.p--
",'0 o
""
II
III
OHM-eM
J..~"c'
V
m~f
-;; " "'I:~U
V,
/~
PERMEABILITY- GAUSSES/OERSTEDS
1132"Cu
,.,
GENERAL
PREDICTION
!i=>
I-
1Il.~~
./
.~:: ~r::~!~~::T~~~AaILIT~
Z
o
l-
1= o;;;r;:~
~ f-tll
";!;
~
II4"F'
~.~
"'X.
rn=I.~89.IO-4~
-
ATTUndlh= 20 LOG,(ei}
.(I31Cu
--
,,'
r:/'
,~V
V ,
V
0/
.D40AI
V
~
1
,.,1--'
LOW PERM STEEL
p.=100 .=0.447'"
,
/'"
I
,
"
;;i--'
10
FIGURE 17.10
-
-COPPER m=O.151{f
J..-"
,
1..-......
•
100
I
,
•
IK
•
FREQUENCY. Hz
10k
3
111111
I 111111
•
lOOK
I I I
3
1•
1M
PENETRATION LOSS (ATTENUATION) OF DIFFERENT SHIELDING MATERIALS
The design of the shielded compartment can be equally as important. Figure 17.10 shows the effectiveness of different materials and
thicknesses in reducing the radiated magnetic field.
496
RADIO FREQUENCY INTERFERENCE AND INTERACTION OF THYRISTORS
17.2.5 Zero Voltage Switching
As we have seen, the RF noise contribution of thyristors is primarily due to a sudden step in current as the thyristor switches. In
some applications, particularly in electric heating, satisfactory control
may be obtained by turning the thyristors on at line voltage zeros,
giving only complete half cycles of current to the load. By eliminating
the sudden steps of voltage, the RF noise contribution is brought to an
absolute minimum. This eliminates the need for R-F filter components,
which, for a large heating load, can become quite large and costly.
For details on this type of circuit, see Chapter 11.
17.3 INTERACTION
In some instances the thyristor system acts as a "receiver" of voltage transients generated elsewhere in the circuit. These transients act
either (or both) on the thyristor trigger circuit or directly on the anode
of the thyristor in the main power circuit. Interaction will cause the
thyristor system acted upon to completely or partially follow, or track,
another thyristor system. Also, various types of partial turn on, depending on the nature of the trigger circuit, have been known to arise.
Elimination of interaction phenomena must take total system layout
into consideration. Section 17.6 gives some general design practices
which should be followed to minimize possible sources of interaction.
Beyond good design practice in the system as well as in triggering
circuits very specific steps for decoupling can be taken as outlined for
UJT circuits in Section 17.4. In general it should be noted that the
suppression of RFI emanations also serves to minimize susceptibility.
17.3.1 Interaction Acting on Anode Circuit
When a thyristor circuit is acted upon with its gate circuit disconnected (open gate or terminated per specification bulletin) the nature
of the interaction is usually attributable to a rate of rise of forward
voltage (dv/dt) phenomenon. When energizing the circuit, such as by
a contactor or circuit breaker, applicable dv/dt specifications for the
device must be met. This subject is discussed in detail in Chapters
3 and 5. Once the circuit is energized the thyristor will sometimes
respond to high frequencies superposed on the anode supply voltage.
For example, a I-megacycle oscillation having a peak amplitude of
10 volts has an initial rate of rise in the order of 60 volts per microsecond. Applicable specifications for the thyristor must meet this condition or steps should be taken to attenuate the rate of rise of voltage.
Due to the nature of anode circuit interaction a thyrisor will
rarely track another circuit over the full control range of phase control.
Usually, it will tend to lock in over a very limited range near the top
of the applied anode voltage half cycle where the dv I dt is greatest.
The best means of suppressing this type of interaction is to select a
device with increased dv/dt withstand capability, to increase dv/dt
withstand capability by means of negative gate bias, or, conversely, to
497
SCR MANUAL
reduce the rate of rise of positive anode voltage by suitable circuit
means. The effect of negative gate bias on SCR dvI dt withstand capability and dv/dt suppression circuitry is discussed in Section 3.11.
Often a combination of these steps yields the desired results. In addition, of course, good circuit layout and system practices should be
observed as outlined in Section 17.6.
17.3.2 Interaction' Acting on the Trigger Circuit
There are basically two cases to distinguish here:
1. The trigger circuit is acted upon from the supply line
directly;
2. The trigger circuit is acted upon from the thyristor gate
circuit.
Both of these mechanisms may cause the trigger circuit to fire
prematurely, giving rise either to spurious triggering or complete or
partial tracking of the thyristors in the circuit. The response of the
trigger circuit to incoming transients will determine the degree of interaction, if any. There are no general rules for every type of trigger circuit. However, in the design of a trigger circuit it is well to take the
possibility of interaction into account. The designer will be in the best
position to assess the transient susceptibility and stability of his circuit.
When using the unijunction transistor trigger circuit there are a
few relatively simple steps that can be taken to decouple these circuits
against both supply voltage and gate circuit transients. These methods
are outlined in the following two sections.
17.4 DECOUPLING THE UJT TRIGGER CIRCUIT
AGAINST SUPPLY TRANSIENTS
Depending on the nature of the particular circuit conditions, either
one or a combination of the following will give effective decoupling
against line voltage transients acting on the unijunction transistor
trigger circuit:
1. Use of control (isolation) transformer with a properly grounded
shield between primary and secondary or an RF filter across
its secondary, if necessary;
2. Use of "boot strap" capacitor between base two and the emitter of the unijunction transistor;
3. Use of a Thyrector diode connected across the supply to the
unijunction circuit.
The value of the "boot strap" capacitor C 1 should be chosen so
that the voltage divider ratio of C 1 and C 2 in Figure 17.11(A) is
approximately equal to the intrinsic standoff ratio of the UJT, or:
C1
If this condition is met, positive or negative transients on the unijunction supply voltage will not trigger the UJT.
498
RADIO FREQUENCY INTERFERENCE AND INTERACTION OF THYRISTORS
+0-..._----,
+O----,~----,
QI
(A) POWER SUPPLY
TRANSIENTS
FIGURE 17.11
(8) TRANSIENTS FROM
SCR GATE CIRCUITS
CIRCUITS FOR ELIMINATION OF ERRATIC FIRING FROM VOLTAGE
TRANSIENTS IN UJT CIRCUITS
17.5 DECOUPLING UJT CIRCUITS
AGAINST SCR GATE TRANSIENTS
Negative voltage transients appearing between the gate and
cathode of the SCR's when transmitted to the UJT can cause erratic
triggering. When transformer coupling is used, these transients can be
eliminated by using a diode bridge in the gate circuit of the SCR as
shown in Figure 17.11(B). Negative transients often arise in SCR gate
circuits in forced-commutated circuits (see Chapter 5) and under certain conditions in AC phase control circuits.
17.6 GOOD DESIGN PRACTICES TO MINIMIZE
SOURCES OF SCR INTERACTION
Radio frequency interference and interaction are both total system phenomena and no one step is necessarily the most eHective in
attaining the desired level of suppression. A combination of good
system design practices, good circuit layout, good equipment layout,
and, if necessary, a small amount of circuit filtering, as was outlined
above, will suppress RFI to acceptable levels and eliminate various
types of interaction phenomena.
When the following system considerations are met it is often
unnecessary to take additional specific steps to filter trigger or anode
circuits (Section 17.2) or use negative gate bias and dv/dt suppression
circuitry (Section 3.11):
499
SCR MANUAL
1. Operate parallel and potentially interacting thyristor circuits
from a stiff (low reactance) supply line;
2. If supply line is soft (high reactance), consider using separate
transformers to feed the parallel branch circuits; each transformer should be rated no more than the required rating of
the branch circuit load;
3. Avoid purely resistive loads operating from stiff lines-they
give highest rates of current rise on switching;
4. Keep load moderately. inductive-limiting rate of current rise
on switching is in the direction of attenuating RFI and minimizing the possibility of interaction;
5. Keep both leads of a power circuit wiring run together-avoid
loops that encircle sensitive control circuitry;
6. Arrange magnetic components so as to avoid interacting stray
fields.
7. Use twisted pair or shield wires for control power wiring and
gate driver wiring.
17.7 E.M.I. STANDAIDS AND RESTRICTIONS
1. American National Standards Institute, Inc., 1430 Broadway, New
York, N. Y. 10018.
C63.4-1963 "Radio-Noise Voltage and Radio-Noise Field Strength"
C63.2-1963 "Radio-Noise and Field Strength Meters 0.015 to 30
Megacycles/Sec."
2. National Electrical Manufacturers Association, 155 East 44th Street,
New York, N. Y. 10017
WD2-1970 "Semiconductor Dimmers for Incandescent Lamps"
3. Department of Defense, Washington, D. C. 20360
MIL-STD-461A "Electromagnetic Interference Characteristics
Requirements for Equipment"
MIL-STD-462 "Electromagnetic Interference Characteristics,
Measurement of"
MIL-STD-463 "Definitions and System of Units, Electromagnetic
Interference Technology"
4. Federal Communications Commission
Title 47 CFT, Chapter I, Part 2, Sub-part I, Paragraphs 2.801 to
.
2.813.
Printed in Federal Register, Vol. 35, No. 100, May 22, 1970, pages
7898-7899.
Vol. II of the FCC Rules and Regulations includes the following
up-dated parts of interest:
Part 2, Frequency Allocations and Radio Treaty Matters: General
Rules & Regulations
Part 15, Radio Frequency Devices
Part 18, Industrial, Scientific and Medical Equipment
(Can be ordered from the Superintendent of Documents, Government Printing Office, Washington, D.C. 20402. Substitute pages,
incorporating amendments, will be mailed to all purchasers of
this volume.)
500
RADIO FREQUENCY INTERFERENCE AND INTERACTION OF THYR1STORS
5. Comite International Special des Perturbations Radioelectriques,
Verband Deutscher Elektrotechniker-Verlag, Berlin 12, Bismarkstrabe 33, VDE-0872 through VDE-0877.
REFERENCES
1. "Application of Fast Recovery Rectifiers," J. H. Galloway, General
Electric Company, Auburn, N. Y. Application Note 200.38.*
2. "Frequency Analysis Modulation and Noise," book, Stanford
Goldman, McGraw Hill Inc., New York, 1948.
3. "Radio Frequency Interference," a series of editorial features
appearing in "Electronic Industries," 1960-1961.
4. "Electrical System Transients and Sensitive Circuit Control," T. B.
Owen, AlEE Applications and Industry, November 1960.
5. "One-Point Ground System With RF Shielding and Filtering,"
R. A. Varone, AlEE Conference Paper 60-1067, Fall General
Meeting, August 1960.
6. "How to Locate and Eliminate Radio and TV Interference," book,
R. D. Rowe, John F. Rider, Inc., New York, 1954.
7. "Transmitter-Receiver Pairs in EMI AnalYSiS," J. A. Vogelman,
Electro-Technology, November 1964.
8. "Electromagnetic-Interference Control," Norbert J. Sladek, ElectroTechnology, November 1966.
9. "Signal Conditioning," Gould Brush, Application Booklet No. 101,
Gould Inc., Cleveland, Ohio.
10. "Interference Control Techniques," Sprague Electric Company
Staff, Technical Paper No. TP62-1, Sprague Electric Company,
North Adams, Mass.
11. "Radio Frequency Interference," Onan Staff, Onan Division of
Studebaker Corp., Minneapolis, Minn.
12. "Interaction Between SCR Drives," Ben Stahl, IEEE Transactions
on Industry and General Applications, Vol. IGA-4, No.6, November/December 1968.
*Refer to Chapter 23 for availability and ordering information.
501
SCR MANUAL
NOTES
502
MOUNTING & COOLING THE POWER SEMICONOUCTOR
18
MOUNTING & COOLING THE POWER
SEMICONDUCTOR
Successful application of SCR's depends to a great extent on
adequate cooling of these devices. If junction temperature of an SCR
rises high enough, permanent damage may occur in its characteristics
and the device may fail by thermal runaway and melting. Circuits may
fail before thermal runaway or melting in the SCR occurs since insufficient cooling can reduce the forward breakover voltage, increase SCR
tum-off time, moving these and other SCR characteristics outside
specifications sufficiently to induce circuit maHunction. For these
reasons, all SCR's and rectifier diodes are designed with some type of
heat transfer mechanism to dissipate internal heat losses.
Mounting surfaces are generally an integral part of an SCR's heat
transfer path. Proper mounting is always needed for successful SCR
cooling. Thus cooling and mounting the SCR are part of the same
problem and must be treated together.
18.1
LEAD~MOUNTED
SCR's
For small lead-mounted SCR's like the C3, C5, C6, C7, C8 and
C103 series, and some configurations of the C106, C107 and C122
(see Figure 18.1), cooling is maintained by radiation and convection
from the surface of the case and by thermal conduction down the leads.
Several good common sense practices for minimizing the SCR
temperature should be used whenever possible. Minimum lead length
to the terminal board, socket, .or printed board permits the mounting
points to assist in the cooling of the SCR most efficiently. Other heat
dissipating elements such as power resistors should not be connected
directly to the SCR leads where avoidable. Also, high temperature
devices like lamps, power transformers, and resistors should be shielded
from radiating their heat directly on the SCR case. To increase heat
-dissipation of the standard TO-5 case, clip-on transistor radiators are
available from a number of commercial vendors.
Several of the General Electric lead mounted SCR's in the TO-5
case are also available on a power transistor type of base for attachment by clamping screw or the like to a heatsink or chassis. Directions
for mounting these devices are given on the specification sheet for that
type ofSCR.
FIGURE 18.1
LEAD MOUNTED SCR's
503
SCR MANUAL
18.2 MOUNTING SCR's TO HEAT EXCHANGERS
The importance of proper SCR mOllllting can be seen from Figure
18.2.The electrical circuit analog for an SCR's thermal path shows the
mounting interface, Recs, to be a series limiting factor to the How of
thermal power (heat) from the jllllction to the ambient. Attention is
not focused on ReJc in this chapter since it is beyond the control of
the equipment designer. Mention is made of it regarding selection of the
proper value from the SCR specification sheet where an SCR has a
multivalued Rem. Somewhat like a chain, a series circuit is limited by
its weakest link, i.e., highest resistance component. In order to prevent
Recs from becoming the weak link, general instructions and guide
lines for the proper mounting of the various types of SCR packages
will be discussed first. Following the general guide lines specific sections deal with considerations peculiar to each package type.
i------T,;-p~----l
(~uNeTION)'
I
:
I
seR
PACKAGE I
I
RSJC:
I
L______ _______
I
I
(CASE'
Te
I
~
R8CS
(EXCHANGER)
(T AMBIENT)
Ts
-=
PO" TOTAL DEVICE POWER DISSIPATION
Te -TA • Po ("8es + "8SA'
FIGURE 18.2 EQUIVALENT THERMAL RESISTANCE NETWORK ANALOG FOR A
POWER SEMICONDUCTOR COOLING PATH
18.2.1 Case to Heat Exchanger Interface Considerations
The interface formed between the SCR package case and the heat
exchanger can take many forms. The corresponding values of case to
heat exchanger thermal resistance will vary greatly depending upon
the given interface conditions.
Figure 18.3 illustrates the effect of metal surface conditions on
interface performance. The exchanger surface distortion as well as
smoothness of surface finish is exaggerated to show its effects.
F
THERMAL
I
EXCHANGER
l
GREASE~
INSULATOR
OPTIONAL
-i6~~~~!
EXCHANGER
~~~~=~r--INSULAtOR
THYRISTOR
CASE
(I) Before Mllllting Farce
(b) After Mlunting Farce Applied
FIGURE 18.3 EFFECT OF INTERFACE SURFACE AND FlATNESS CONDITIONS
ON THERMAL CONDUCTIVITY
504
MOUNTING & COOLING THE POWER SEMICONDUCTOR
After force is applied to the joint the surfaces are forced together
at the points of contact. The net contact area is then a function of contact metal ductility, surface finish, flatness and net force applied. In
addition a thermal grease is shown which serves to fill in the voids
left by the valleys due to poor surface finish. Note from Figure 18.3(b)
the large void in the top contact area due to the poor Hatness of the
sink as compared to the bottom interface where the insulator and thyristor case are shown to be Hat. The insulator's thickness and thermal
conductivity further adds to the interface resistance.
18.2.1.1 Exchanger Surface Preparation
The surface under the semiconductor contact surface should be
Hat to within 0.001 inch per inch and have a surface finish of 63 microinches or less for all stud and tab mounted devices. For press paks
exchanger surface should be Hat to within 0.0005 inch per inch and
have a surface finish of 32 micro-inches or less.
Before final assembly, the semiconductor case surface should be
checked for removal of all burrs or peened-over corners that may have
occurred during shipping and subsequent handling and that would
otherwise cause reduced heat transfer across the surfaces.
Most heat exchanger surfaces have some treatment to aid radiation
heat transfer and give corrosion protection. Copper fins are plated,
painted, or ebnoled. Aluminum fins are generally painted or anodized.
The heat exchanger surface under the semiconductor contact surface
must be free of paint, anodization, or ebnol to give minimum contact
thermal resistance. While plating in this area does not have to be
removed, excessive oxides should be removed whenever the exchanger
surface has been exposed to the ambient air for more than sixty minutes
after machining.
Oxide removal prior to assembly may be accomplished by polishing the mounting surface with No. 000 fine steel wool and silicone oil.
Following the polishing the surface should be wiped clean with a lint
free paper towel. As a final step a thin layer of thermal grease or oil
should be applied.
Applications where a moist or corrosive atmosphere are expected,
galvanic action between aluminum and the copper SCR case may lead
to gradual deterioration of the joint, and an increase in thermal resistance. A good nickel (ALSTAN 70 Process), silver or cadmium plate
over the copper case as provided on General Electric SCR's, combined
with the use of a corrosion inhibitor, such as Burndy-Penetrox A; Alcoa
No.2, Dow Corning DC 19 or Penn-Union Cual-Aid, minimizes corrosion at this joint.
18.2.1.2 Interface Thermal Grease
Thermal greases serve two functions. They serve to resist corrosion
and secondly they enhance the interface substantially by filling in the
voids with a more thermally conductive material than air as shown in
Figure 18.3. Note the decreased thermal resistance of interfaces having
grease over those without grease shown in Table 18.1. Thermally conductive greases and oils come with and without metal fillers. These
505
SCR MANUAL
susp~nded metal fillers generally serve to enhance the joint's thermal
properties. over the non-filled greases but not to a substantial value:
They have the disadvantage of indenting the mounting surface slightly
requiring a careful refinishing of the surfaces with 400 or 600 grit sandpaper should disassembly or reassembly become necessary. Table 18.2
provides an application guide to the many oils and greases available to
the user. The chief advantage of the oil over the grease liesin the better
control of film thickness possible with oil. The user can specify one or
two drops from an eye dropper. Excessive grease or oil can be detrimental to interface thermal resistance.
Stud Size
10-32
X"-28
X"-28
%"-24
Y2"-20
%"-16
%"-16
Flat Based
Hex Size
Across
Flats or
Flat
Base Dia.
'KI'
~/'
lJ{l'
IJ{/'
IJ{/'
IX"
1%"
1%"
Case-Exchanger Thermal Resistance
~ Recs - °C/Watt--')
With Thermal Grease
Dry
Min. Nom. Max. Min. Nom. Max.
.09
.07
.05
.02
.02
.025
.015
.01
.3
.25
.15
.06
.065
.08
.04
.025
.8
.6
.4
.15
.2
.2
.10
.07
.2
.15
.10
.05
.05
.06
.03
.5
.4
.25
.1
.12
.15
.07
1.2
.9
.6
.25
.3
.35
.15
Stud Insulated With 5 Mil Mica Washer
'KI'
10-32
X"-28
X"-28
~"
WI'
1.2
.9
.7
2.5
2.0
1.5
4.5
3.5
2.5
PRESS PAK SINGLE INTERFACE - LUBRICATED
PressPak
Interface Dia.
%"
1"
IX"
IX"
TABLE 18.1
506
Nominal Clamp
~ Recs - °C/W --')
Force
Minimum Nominal Maximum
800
2300
2300
4000
0.04
0.02
0.015
0.014
0.06
0.03
0.022
0.02
INTERFACE CASE TO EXCHANGER THERMAL RESISTANCES
0.20
0.10
0.08
0.07
MOUNTING & COOLING THE POWER SEMICONDUCTOR
-APPLICATIONCorrosive & High Moisture
Ambient Environment
Plated Heat Exchangers Necessary
Silicone Oil
Silicone Grease
Dow Corning
DC 703
Dow Corning
DC 3,4, 340
and 640
General Electric General Electric
SF 1017
G623
Dry, Pollution Free Ambient
Environment
A) Heat Exchanger unplated:
Dow Corning DC19
Burndy Penetrox* A
Alcoa #2
B) Heat Exchanger plated:
Grease or oil not critical
*Contains Filler Particles.
Additional oils and greases available from all major heat exchanger
suppliers. The above thermal oils and greases have been tested and
stressed by thermal cyclic life testing.
TABLE 18.2
THERMAL COMPO UNO APPLICATION TABLE
The values of Recs given in Table 18.2 under the nominal heading are easily achieved by following the recommendations listed above
and on the thyristor data sheets. If one or more surfaces are not per
recommendations, values of Recs can easily reach the maximum values
indicated and under extreme conditions exceed given values where
torque or force applied is grossly misapplied. Minimum values are
achievable under tightly controlled assembly conditions. It is not recommended that minimum values be used for design purposes unless
quality sampling audits are made to ensure conformance to design
values.
18.2.1.3 Electrical Isolation Case to Heat Exchanger
In some applications it is desirable to electrically insulate the semiconductor case from the heat exchanger. Hardware kits for this purpose are available for stud-mounted semiconductors with machine
threads in the low and medium power ratings. These kits generally
employ a .003 to .005 inch thick piece of mica or bonded fiberglass to
electrically isolate the two surfaces, yet provide a thermal path between
the surfaces. As evidenced by the data in Table 18.2, the thermal
resistance of the joint may be raised as much as ten times by use of
this insulation. As in the direct metal-to-metal joint, some improvement
in thermal resistance can be made by using grease on each side of
the mica.
Tests using beryllium oxide (99 per cent) for electrical insulation
have shown this material to be excellent in heat transfer. Insulating a
semiconductor with a %-20 stud, using BeO (99 percent) washers
(1.00 inch OD x .52 inch ID x .125 thk) gave a stud-fin contact thermal
resistance of 0.14°C/watt. Applying thermal grease to all contact surfaces decreased that thermal resistance to 0.1°C/watt.
507
SCR MANUAL
Beryllium oxide discs are also available with one or both sides
metalized. With this metallization it is possible to solder the semiconductor case, the heat exchanger, or both to the BeD disc. This technique is particularly useful with the flat-bottom press fit package.
Figure 18.4 shows the drastic improvement in case-to-sink thermal
impedance using the metalized BeD and solder technique.
ISOLATION METHOD RS
cs
_oC/W
~-:-----"
~
",SOLDER
B.O~EPO~
_____ O'77
980
~
n
U
.
NOTES'
SOLDER
.
.
- - - - -- 0.30
I. SOLDER - 60/40 (Ph/Sn)
2.EPOXY- HYSOL #A7-4322
(CURE f·50 0 e.7 HOURS)
FIGURE 18.4 MOUNTING THE PRESS FIT PACKAGE WITH BERYLLIUM OXIDE INSULATION
Low Power stud thyristors are available from General Electric
with integral beryllium oxide washers as shown in Figure 18.5. The
additional thermal resistance,' ReJc , due to the beryllium isolation is
O.3°C/watt for the 1/4-28 stud shown in Figure 18.5.
~PRESS-FIT
ISOLATED
STUD
STUD
FIGURE 18.5
508
EXAMPLE OF AVAILABLE BERYLLIUM OXIDE ISOLATED STUD THYRISTORS
MOUNTING & COOLING THE POWER SEMICONDUCTOR
WARNING: Beryllium oxide discs and/or products incorporating
beryllium oxide ceramics should be handled with care. Do not crush,
grind, or abrade these portions of the thyristors because the dust resulting from such action is hazardous if inhaled.
Beryllium oxide washers in large, formed sizes and small quantities are basically somewhat expensive items. However, careful consideration should be given to the over-all economics before using any
other material when electrical insulation is required. Several standard
washer sizes are available from companies such as National Beryllia
Corporation, Frenchtown Porcelain, or Brush Beryllium Company.
Another method used for insulating the semiconductor case from
the heat exchanger is to directly solder the device to a small metal
plate and then insulate that from the heat exchanger. Figure 18.6
shows this approach used with the flat-bottomed press-fit package.
The SCR is soldered to a flat metal plate (say 3 or 4 square inches in
area). Soldering must be accomplished below 200°C; a 60-40 (Pb-Sn)
solder can be employed at about 180°C. A solder which looks quite
promising for this application is "Alloy 82" from Alloys Unlimited Inc.
This is a lead-tin-indium (37%-37%-25) perform of 468 mils diameter
and 9.5 mils thickness. Soldering with this alloy can be accomplished
at 150°C.
SOLDER
047- BASE
METAL PLATE
DIAMETER--..
rAPE
HEAT
EXCHANGER
OR
CHASSIS
FIGURE 18.6 MOUNTING THE PRESHIT PACKAGE WITH TAPE INSULATION·
The SCR-flatplate assembly is then mounted to the heat exchanger
by means of an epoxy coated, mylar tape. The tape recommended is
Scotch* Brand #75. This is a one mil mylar tape with about 2 mils of
epoxy (when cured) on both sides. The epoxy cures at 121°C in three
hours.
The advantage of the above outlined mounting technique is that
the operating heat generated at the device junction is first spread out
over a large physical area before it tries to traverse the insulating
medium. This reduces the total thermal resistance of that insulation.
Other techniques employing the direct use of epoxy adhesives,
epoxies with a filler, and the direct use of insulating tape have been
successfully employed. 4 Since in most cases a rather high price in current rating is paid in thermal resistance for insulated mounting, such
mounting is not recommended for high power SCR's.
For comparison of the tape insulation method with the beryllium
oxide method of Figure 18.4 a conservative case to exchanger value
of 1.2°C/watt has been measured using the following geometry and
a press fit package as shown in Figure 18.6. 2" x 1 %" x .050" copper
spreader plate with a Scotch* Brand #75, one mil mylar tape, with
2 mil epoxy layer (when cured) on each side.
Figure 18.7 illustrates the design trade-off factors available to the
designer regarding heat spreader plate size and plate thickness.
*Trademark of 3M Corp.
509
SCR MANUAL
2.0
I..
.6
~.4
~
.3
.2
.15
,
~
t\.
~
~
1\
III
COPPER
SPREADER THICKNESS-INCHES
I.
III
.062'
III
:\~'2'
~~,- .2.0
~
~ .~:, IN'"
III
-
NOTE I: (SOLID tuRVES)
1.062'~ t~ ..
CONSERVATIVE AIR FILM
FACTOR INCLUDED.
-
NOTE 2: tDASH CURVES}
ASSUMES IDEAL EPOKY
SPREADER PlATE AND
EXCHANGER BOND.
i-"~i'" '\
-~"
"
.250 r.12'
FOR PRESS FIT THYRISTOR
BASE CONTACT DIAMETER 0
-
OF 0.47 INCHES .
I
II
SPR(ADER THICKNESS -INCHES
.10
1.5
2
61015<1)
3040
SPREADER PLATE ~.D-INCHES
FIGURE 18.7
EFFECT OF SPREADER PLATE AREA & THICKftESS ON CASE TO EXCHANGER R.
The curves are based on the assumption that the heat spreader
plate acts like a heat exchanger fin with cooling on one side only.
Thus the nomograph of Figure 18.31 is applicable. The design example below details the use of the nomograph for this application. Figure
18.7 is applicable for square plates as well. Using the conversion factor
D = 1.128E, where D is the spreader plate diameter as given in the
figure and E is the length of the side of a square plate.
Example heat spreading plate calculation
Given: - Copper spreading plate 2" x 2" square and 0.10" thick
- Thyristor press fit base diameter d = 0.59"
- Insulation material
1 mil mylar
2 mils epoxy each side (cured)
Problem: Determine case to exchanger thermal resistance Recs
Solution:
1st step: Find total heat transfer coefficient, h. For mylar
and epoxy:
.
watt in
.
h m 6 x= 0.004 ~C where 6x = film thICkness,
m
0.005 inches
0.004 ~~t: in
Then h m = 0.005 i:n C = 0.8 (watt/in2 DC)
and h = 1/
(...!..
+ ..!.)
h
h
m
f
where hr is due to air film in epoxy voids. l/h f is conservatively found to be equal to 1.25.
Then h = 1/0.8 ~ 1.25 = 0.4. Since h has the units of thermal conduction/area, 1/h has the units of thermal resistance
x area.
510
MOUNTING & COOLING THE POWER SEMICONDUCTOR
2nd step: Determine fin effectiveness from nomogram of
Figure 18.31.
D = 1.128E where E is the square fin size
D
1.128 x 2"
= 2.256"
D-d
b
-2(2.256 - 0.47)/2
=
=
D/d
=
= 0.89
= 2.256/0.47
=4.8
8 = 0.100 inch plate thickness
Using the nomogram:
Note: the h value is halved. Since the nomogram was designed
for both sides of the fin to transfer heat and only one side of the
plate is assumed to be conducting heat.
Thus for h/2
.2 watt/in2 °C and thickness 8
0.100 inch,
(% = .6.
0.89 and (%
.6, a line is extended to the graph
Through b
where Did = 4.8, 'YJ is found to be 0.83.
=
=
=
=
3rd step: Determine Recs
1
1
Recs ='-- = - - - - - - - - - 7J A h
0 .83 X 22'ill2 X 0.4. watt
20C
ill
-
1
1.33
= 0.75 °C/watt
18.2.2 Mounting the Power Tab
Figure 18.8 shows various configurations of the power tab package. Some configurations of the package are provided with an anode
tab for mounting directly to an appropriate heat exchanger. Because
of this unique package design, it can be mounted in a variety of
methods, depending upon the heat exchanger requirements and the
circuit packaging methods.
As a service to its customers, the General Electric Company provides a lead and tab shaping capability. Any of the derived types
shown in Figure 18.8 are available.
The tab and the leads will bend easily, either perpendicular to
the Hat or to any angle, and may also be bent, if desired, immediately
next to the plastic case. For sharp angle bends (90° or larger), a lead
should be bent only once since repeated bending will fatigue and break
the lead. Bending in other directions may be performed as long as the
lead is held firmly between the case and the bend, so that the strain
on the lead is not transmitted to the plastic case.
511
SCR MANUAL
The mounting tab may also be bent or formed into any convenient
shape so long as it is held firmly between the plastic case and the area
to be formed or bent. Without this precaution, bending may fracture
the plastic case and permanently damage the unit.
When used as a lead mounted device, without heat exchanger,
thermal characteristics are available from the device data sheet in the
form of ambient temperature vs on-state current curves for all four
types.
In-line sockets to accommodate the 2 or 3 Hat leads of the power
tab package are available for printed circuit board or chassis mounting.
These sockets, No. 77-115 (press-fit) or No. 77-116 (Hange), may be
obtained from the Connector Division, Amphenol Corp., 1830 South
54th Avenue, Chicago, Dlinois 60650
DERIVED TYPES
(THE TYPES sttOWN IIIELOW ARE DERIVED
rJlR~=ll~~::~S~~~'":l~~T.ATED
BASIC TYPI!:S
=.r:°'%W~U'G
IIWRIGHTOR FLAT!
RIVET OR
5CMWIIOUI!ITlNG
TO FLAT SURfACE
~
i
.... '
t
I
~ ....
'.'
ft
.... '
"
.....
1
.... "
."."
"
1
."."
•
~......
FIGURE 1&.8 POWER TAB PACKAGE CONFIGURATIONS
Insulated hardware kits are available from General Electric. Kit
details are given on power tab device specification sheets.
When mounting the power tab package to a heat exchanger the
tab or "case" to exchanger thermal resistance Racs is a function of
mounting method, Table 18.3 illustrates the various combinations of
mounting methods with· accompanying values of Racs. The reference
point for tab temperature measurements is illustrated by Figure 18.9.
i
"CASE" TEMP.
REF. POINT
TABOR
(Tel
f
-.IillS"
T
FIGURE 18.9 TAB OR "CASE" TEMPERATURE REFERENCE POINT FOR POWER TAB PACKAGE
512
MOUNTING ILLUSTRATION
~
====1C==:J
~-.-"""'"
: /"'"'""""""
CJ'~
HFH_~.c~,,'(~_rR
'~':l:,!:'
==U'
:;~i~~~~S ---'
~_._"m,"
~
"""
:J- ~~~~:t:LW~1~A~;
CJ'[::=J
THERMAL
GREASE
None
None
None
.003" Mica
.003" Mica
.002" Mylar
Tape (3)
.002" Mylar
Tape
.006" Black
Elect. Tape
.006" Black
Elect. Tape
Yes
None
None
Yes
FASTENER
USED
Screw(l) I
Rivet(2)
ScrewIRivet
60/40 Solder
ScrewIRivet
ScrewIRivet
None
ScrewIRivet
10.3
Yes
Screw IRivet
5.7
None
ScrewIRivet
12.5
Yes
Screw IRivet
10.3
Notes:
-
TABLE 18.3
2.0
.25
9.15
3.75
s::
o
c:
:z
=!
:z
C')
QO
C')
o
6-32 screW torqued to 5-6 in-Ibs.
(2) Use good quality ~Y' diameter semi-tubular rivet and eyelet
of brass backed .250" OD washers under both ends. Top
washer may be omitted if rivet head is larger than 0.250"
diameter. Hydraulic or pneumatic force should be used for
rivet pressure application. Follow rivet manufacturer's instructions for force level values.
(3) Tapes greased on non-stick surfaces only.
(1)
U1
w
NOMINAL
Reos
°C/WATT
5.25
INSULATING
MATERIAL
None
NOMINAL Roes FOR POWER TAB PACKAGES
o
:z
C')
!::
-t
::z:
,...,
~
,...,
::c
en
,...,
s::
c=;
o
:z
c
c:
~
::c
SCR MANUAL
18.2.3 Mounting the Power Pac Package (10-220)
i
2• - .. (Tel
POINT A
POINT
e
ITLI
Use of GE Series 2500 Force Gauge
To Calibrate Force Gauge:
If the gauge is su~ected of being out of calibration due to wear or
damage, check it on a Hat surface as shown below.
If the points are not 0.300 ± .010 apart, calibrate the gauge by filing
the bottom contact points.
(b> Calibration of GE Series 1000 and
2500 Force Gauges
FIGURE 18.23
USING THE PRESS PAK CLAMP FORCE INDICATOR GAUGE
525
SCR MANUAL
c) Force spreading. The force should be transmitted through the
clamp insulator by means of a pressure pad to prevent possible
fracturing of the insulator and cold flow due to application of
excessive compressive forces in the insulating material (see
Figure 18.22).
d) Insulator dimensions. The insulator should provide creep and
strike distances equal to or greater than the press pak it is
clamping.
e) Provisions for maintaining assembly rigidity should be part
of the clamping system. Figure 18.21(b) illustrates the uneven
contact surface resulting from too weak of an exchanger on
the bottom surface of the press pak. The GE clamps are supplied with an optional stiffening brace shown in Figure 18.22
(Note 1) to preclude this problem.
f) Temperature Limitations. All insulating materials and springs
have temperature limitations.
Component limitations should be given. This is most apt to be a
problem when cooling rectifier diodes where case temperatures greater
than 100°C are the norm. The series 1000 and 2500 press pak clamps
are capable of operating at insulator temperatures up to 125°C and
spring temperatures up to 110°C. Clamp temperature can be reduced
by inserting ceramic or other poor thermal conductivity material
between the clamp insulator body and the heat source. The material
should be mechanically stable with time at temperature to insure maintenance of clamp force levels by preventing spring relaxation. Rectifier
diodes operated at rated T;r and single side cooled will always require
a thermal insulating member between clamp and cell when the clamp
insulator is opposite the heat exchanger.
Press pak semiconductors may be mounted using other than GE
clamps but attention of force requirements listed on device data sheets
(generally 800 lbs for :!h" press paks and 2000 lbs for the 1" press pak
except the 600 series which requires 4000 lbs per cell) must be adhered
to in addition to the above requirements. Table 18.5 gives the basic
data for GE power pak clamps.
Data
Clamp Sheet
No.
No.
(1)
700-900
Dimensions A, B, C and D refer to Figure 18.22.
TABLE 18.5
526
Mounting
Hole
Diameter
Inches
.885 x .855 2.115-2.135 0.516 ± .005
x 2.965
170.49 2200-2400 1.000 x 1.270 3.095-3.105 0.875 ± .005
x 4.520
1000 170.48
2500
Force
Range
Pounds
Center Line
Mounting
Insulator
Hole
Dimension
Inches-Max. Dimension
ABC(l)
Min.-Max.
Inches D(l)
PRESS PAK CLAMP DATA
MOUNTING & COOLING THE POWER SEMICONDUCTOR
Detailed mounting instructions concerning clamp tightening and
available bolt lengths are found on the clamp data sheets. Press pak
interface thermal resistance values; case-exchanger are found in
Table lB. 1.
18.2.7.2 Multiple Unit Mounting
18.2.7.2.1 Parallel
The symmetry of the top and bottom surfaces of the press pak
permit a variety of mounting configurations. In some applications it is
desirable to mechanically parallel units. An electrical inverse parallel
connection may be used as in Figure 1B.24(a) or an electrically paralleled connection as in Figure 1B.24(b) may be employed. The devices
must not be clamped between two rigid heat exchangers because of
difference in device height. One rigid heat exchangers may be common
to the units on one side; on the other side either individual exchangers
should be used or the heat exchangers should be flexible enough to permit good contact with each Press Pak surface. Individual clamJls are
needed for each device to provide the specified force. Figure 1B.25(a)
shows a photo of a parallel assembly available from GE as a complete unit.
(b)
(8)
FIGURE 18.24
PARALLEL MOUNTING OF PRESS PAKS
(8) Parallel (Forced Air Cooled)
527
SCR
FIGURE 18.25
EXAMPLES OF MOUNTEO PRESS PAIlS FROM FACTORY
AS STANOARO ASSEMBLIES
18.2.7.2.2 Series
In-line mounting is suitable for many applications. It has the
advantage of employing only one clamp. Many electrical configurations
are possible with in-line mounting. Figure lS.26(a) shows a series string
which could be used for high voltage circuits, or top and bottom terminals could be connected together to produce an inverse parallel pair.
Figures lS.26(b) and lS.26(c) are doubler circuits (cathode common
or anode common); here top and bottom terminals could he tied
together to yield a parallel pair.
··,,
·.,
I
,
(a)
(II)
(c)
FIGURE 18.2& SERIES MOUNTING OF PRESS PAIlS
Figure lS.25(b) illustrates a photo of a series liquid cooled
assembly available from CE. Due to thermal expansion and contraction
care should he taken in seriesing more than two cells in a given assembly. Reference 6 gives detailed calculations for making design decisions
relative to material use and dimensions of a multi-cell assembly.
528
MOUNTING & COOLING THE POWER SEMICONDUCTOR
18.2.7.3 Handling of Press Pak
Although the press pak is a rugged .component, reasonable care in
handling is recommended. Dropping, or other hard jarring, of the
devices can damage the silicon pellet and destroy electrical characteristics. Dents, nicks, or other distortion of the contact surfaces will also
retard the How of heat through the heat exchanger and cause the junction to overheat.
18.2.8 Unit Pak Mounting
The unit pak as shown in Figure IB.27 is intended for one-side
cooling by application to one heat exchanger with two bolts. The unit
pak comes from the factory equipped with both clamp and current
take-off. The standard current take-off is shown, but other configurations are available and are needed should the unit pak be used with
liquid cooled heat exchangers.
FRONT VIEW
FIGURE 18.27
UNIT PAK CLAMP ounlNE
The unit pak is only compatible with C50l, C520, C530, C506,
C507, C50B, C510 SCR's and A500, A540 and A570 rectifier diodes.
18.2.8.1 Preparation of Heat Exchanger
With heat exchanger thickness of .BO", or greater, use tapped
hole for steel Helicoil #llB5-5LN-7Bl. If Helicoil is not used, tap hole
for %6"-IB thread with a minimum of 1" length of bolt engagement in
heat exchanger.
If aluminum, thinner than .BO" but thicker than .375", is used,
the bolts should be secured with the standard %6"-18 steel nut supplied with the Unit Pak
For heat exchanger mounting area thinner than .375" use a hardened back-up plate of steel, approximately 0.5" thick x 1" wide x 3.500"
long.
529
SCR MANUAL
18.2.8.2 Mounting Procedure
1. Check mounting surface for foreign particles, nicks, etc. Wipe
off with lint-free paper towel.
2. Place Unit Pak on greased heat exchangers with end of bolts
in mounting holes. By pressing Unit Pak down to the heat
exchanger the bolts will snap up to align properly.
3. Pull down bolts hand tight with spring leveled. Locate current
take off in desired orientation.
4. Turn each bolt llh turns with a wrench. Adjust level of spring
with an additional 1 to 3 Hats. The spring should now be Hat
and can be checked with a straight edge. Adjust the spring
for Hatness by turning an additional 1 to 3 Hats on each bolt.
18.3 SELECTING A HEAT EXCHANGER
Heat exchanger selection is governed by theSCR package and
environmental constraints. Because the heat exchanger thermal path is
in series with the SCR package thermal resistance, diminishing returns
are reached rapidly when attempts are made to make R esA ·substantially lower than ReJC . For this reason this section is divided into two
sub-sections dependent upon SCR current ratings which in turn determine a particular range of heat exchanger types.
18.3.1 Low to Medium Current SCR's
SCR's in this classification have a ReJC falling in the range of 1 to
lOoC/watt. Consequently heat exchangers typically used have ResA
values of from 1 to 30°C/watt. The design used most often in this
range is the Hat plate or "fin" and variations of it. The Hat fin's chief
advantage is its low cost and design Hexibility. Heat is dissipated from
the fin to the ambient air by both radiation and free or forced convection heat transfer. Fin selection is accomplished by use of design
nomographs as illustrated in the following paragraphs.
18.3.1.1 Designing the Flat Fin Heat Exchanger
18.3.1.1.1 General
Because the mechanisms of radiation and convection heat transfer
are of distinctly different nature, the so-called heat transfer coefficient
530
MOUNTING & COOLING THE POWER SEMICONDUCTOR
Symbol
Definition
A
Surface Area of Fin
c
Thermal Capacity
h
Heat Transfer Coefficient
k
Thermal Conductivity
Length of Fin (in specified direction)
L
q
Rate of Heat Flow
T
Temperature
aT Temperature Difference
Surface Temperature of Heat Exchanger
Ts
TA
Ambient Temperature
V
Air Velocity
£
Radiation Surface Emissivity
."
Fin Effectiveness
ReSA Thermal Resistance (Exchanger to Ambient)
TABLE 18.6
Dimensions
in. 2
watt-sec/lb. °C
watts/in.2 °C
wattsrC-inch
inches
watts
°C
.OC
°C
°C
fUmin.
°C/watt
BASIC THERMAL UNITS
(h) for each effect must be calculated separately and combined with
the fin effectiveness (.,,) to determine the over-all heat transfer coefficient if any degree of confidence is to be placed in the analytical design.
The rate of heat flow, q, from the fin to the ambient air can be
expressed as follows:
hA."aT
(18.4)
q
Correspondingly the fin's thermal resistance can be expressed by
1
ReSA h A."
(18.5)
=
=
where h = total heat transfer coefficient of the fin
A = surface area of the £in
." = fin effectiveness factor
aT
= temperature difference between hottest point on fin
and ambient
'
ReSA
Thermal resistance (exchanger to ambient)
Table 18.6 lists these and other symbols used in the following discussion together with. their dimensions.
A short discussion on each of the major factors in Equation 18.4
will reveal the variables on which they depend. The examples cited
all apply to the same size fin and temperature conditions so that the
reader can compare the relative magnitude of each of the various
mechanisms of heat transfer.
It should be emphasized that while the individual equations are
quite accurate when the conditions on which they are based are fulfilled in detail, the practical heat exchanger design will depart from the
conditions to some extent because of local turbulence in the air due to
mounting hardware and leads, thermal conduction down the electrical
leads and the mounting for the fin, nearby radiant heat sources, chimney cooling effects caused by other heated devices above or below the
cooling fins, etc. Fortunately most of these additional effects enhance
rather than reduce the heat transfer. Therefore, it is common practice
to disregard these fringe effects in the paper design stages except
=
531
SCR MANUAL
where designs are being optimized to a high degree. Even in a highly
optimized design, precisely calculated values may be subjected to
substantial corrections when the design is actually checked in the prototype. The final measure of the effectiveness of the cooling fin will
always be the fin temperature at the case which should never be
allowed to exceed the manufacturer's rating for a given load condition.
18.3.1.1.2 Radiation
For stacked fins with surface emissivity of 0.9 or more and operating up to 200°C, the radiation coefficient (hr ) can be closely approximated by the following equation: *
X10-
hr = 1.47
where
£
(1 - F)
10 €
(Ts ~ T + 273) ;n~a!~ (IS.6)
A
= surface emissivity (see Table IS.7)
= shielding factor due to stacking (F =
0 for single
unstacked fins)
T s = surface temperature of cooling fin (0C)
T A = ambient temperature (OC)
Table IS.7 indicates the wide variation in emissivity for various
surface finishes. In free convection cooled applications, the radiation
component of the total heat transfer is substantial, and it is therefore
desirable to maximize radiation heat transfer by painting or anodizing
the fin surface.
Note that oil paints regardless of color improve surface emissivity
to practically an ideal level (unity).
Figure IS.2S presents Equation IS.6 in the form of a nomogram
which considers the detrimental effects of stacking cooling fins. As
fin spacing is reduced shielding effects become more marked, and radiation heat transfer is reduced.
F
h,
WATTS/IN 2,oc
RADIATION NOMOGRAM
EMISSIVITY -
0
I--
-F::::
a
9
~'"
'"
FIN
~
Ta-GC
MEAN TEMP
1C
SURFACE
at 5
180
00.
ROUND OR
SQUARE FIN
180
006
~ f\. /
I ~ f\. 'Ie-
- -
-
\1\
i\ \
1
\
1
60
.001
\
\1\
I
~
0008
30
20
10
,0005
® ~+t~,--,--,
I
11
liz
I"'-"~'l"-JI""'"",-I
Wi1-1J11J.1hl'
,
20
1 ,LI---,,-"'-----
SMALLER
StOEOR
DIA~I~~~RE~r FIN
I
!l!'1/
v
••
FIGURE 18.28
532
10
LJ'' . . . . . . .
I
i
34 / 6
60
40
.QQ06
I
1
2
200
01
2: I RECTANGULAR
~
S~~~<~i"
I I
(j)
RADIATION NOMOGRAM (EMISSIVITY
= 0.9)
+ T"'Ma)
®
MOUNTING & COOLING THE POWER SEMICONDUCTOR
Surface
Anodized Aluminum
Commercial Aluminum (Polished)
Aluminum Paint
Commercial Copper (Polished)
Oxidized Copper
Rolled Sheet Steel
Air Drying Enamel (any color)
Oil Paints (any color)
Lampblack in Shellac
Varnish
TABLE 18.7
Emissivity (£)
0.7-0.9
0.05
0.27-0.67
0.07
0.70
0.66
0.85-0.91
0.92-0.96
0.95
0.89-0.93
EMISSIVITIES OF COMMON SURFACES
Example of Use of Radiation Nomogram
Given:
- Stack composed of 3" x 3" square cooling fins
- 1" spacing between fins
- ambient temperature = 40°C
- fin surface temperature = 100°C
Problem: Determine coefficient of radiation heat transfer (hr) and total
radiation heat transfer (qr) assuming fin effectiveness = l.
(See Section 18.3.1.1.5)
Solution:
Following the dashed line sequence starting at 1 for the
above conditions, hr
.0024 W/in.2 DC.
qr = hr A AT = (.0024 watts/in. 2 0c) (3 x 3 in.2) (2 sides)
(100 - 40°C) = 2.6 watts per fin.
For single unstacked fins surrounded by 40°C ambient h.
.0054 watts/in. 2 °C by the indicated line on the nomogram.
qr = hr A AT = (.0054) (3 x 3 x 2) (100 - 40) = 5.8 watts
per fin.
=
=
18.3.1.1.3 Free or Natural Convection
For vertical fins surrounded by air at sea level and at surface
temperatures up to 800°C, the free convection heat transfer coefficient
Z
0
80
OJ
OJ
It:
IL
IL
60
u
0
IZ
OJ
U
It:
OJ
......
EFF
T
40
"-
...........
ALTITUDE ON
CTION HEAT
T
F~~D~ZE; 1/2 "
'""
'\
1,\
20
~
10
100
ALTITUDE -THOUSANDS OF FEET
FIGURE 18.30
EFFECT OF ALTITUDE ON FREE CONVECTION HEAT TRANSFER
18.3.1.1.4 Forced Convection
When air is moved over cooling fins by external mechanical means
such as fans or compressors, heat transfer is improved and the convection heat transfer coefficient can be approximated by the following
equation: *
he
= 11.2~ ~ X
10- 4 watts/in. 2 °C
(IS.S)
where V = free stream linear cooling air velocity across fin surface
(ft.lmin.)
length of fin parallel to air How (inches)
L
This equation is based on laminar (non-turbulent) air How which
exists for smooth fin lengths up to L :=; C/V, where C is a constant
given in Table IS.S for various air temperatures. For L > C/V, air How
becomes turbulent and heat transfer is thereby improved. Turbulent
air How and the resultant improvement in heat transfer may be
achieved for shorter L's by physical projections from the fin such as
wiring and the rectifier cell itself. However, turbulence increases the
power requirements of the main ventilating system. Minimum spacing
=
for the above is B
~ ~
inches where B is also a constant given in
Table IS.S.
Air Temperature
25°C
55°C
S5°C
125°C
150°C
B
3.4
3.S
4.1
4.5
4.7
C
37,000
45,000
52,000
63,000
70,000
TABLE 18.8 LAMINAR flOW LIMITATIONS
*This is accurate within 1 % of Equation 7.4S, Reference 2, p. 149 for
air properties up to 250°C.
535
SCR MANUAL
Figure 18.31 presents a nomogram for convenience in solving the
forced convection equation, Equation 18.8 above.
he
FORCED CONVECTION
COEFFICIENT-WATTS
L
-;2':C
LENGTH OF
FIN-INCHES
(PARALLEL TO
AIR FLOW)
.03
V
VELOCITY
OF AIR
LINEAR F.P.M.
--(i)
FIGURE 18.31
FORCED CONVECTION NOMOGRAM
Exallple of Use of Forced Air Convection NOllograll
- 3" x 3" square cooling fin
= 300 linear FPM
40°C
- fin surface temperature = 100°C·
Problem: Determine forced convection heat transfer coefficient (h.,)
and total convection heat transfer (qe) assuming fin effectiveness = 1.
.
Solution: L = 3 inches, V = 300 LFPM
As shown on dashed line on nomogram, he = .011 watt/
in.2°C.
qe = he A AT
(.011 watts/in. 2 0c) (3 X 3 in. 2) (2 sides)
(100 - 40°C) = 11.9 watts.
Given:
- air velocity
- ambient air
=
=
18.3.1.1.5 Fin Effectiveness
For fins of thin material, the temperature of the fin decreases as
distance from the heat source (the SCR) increases due to effects of
surface cooling. Thus calculations of heat transfer, such as those above,
which are based on the assumpti0n that the fin is at a uniformly high
temperature are optimistic and should be corrected for the poorer heat
transfer which exists at the cooler extremities of the fin. The correction
536
MOUNTING & COOLING THE POWER SEMICONDUCTOR
factor which is used is called fin effectiveness (7J). 7J is defined as the
ratio of the heat actually transferred by the fin, to the heat that would
be transferred if the entire fin were at the temperature of the hottest
point on the fin. The hottest spot, of course, is adjacent to the stud of
the SCR. The effectiveness depends on the length, thickness, and shape
of the fin, on the total surface heat transfer coefficient h, and on the
thermal conductivity k of the fin material. As defined in Equation 18.4,
the total actual heat transfer may be calculated by multiplying the fin
effectiveness factor by the total surface heat transfer (determined by
adding the radiation and convection heat transfer as calculated in the
examples above).
a-FIN THICKNESS
2.0
FOR
FOR
TT
'0
-."..
0.1
~~
'.0
.'
0'
.o,
0.2
.0001
4.0
5.0
INCHES
INCHES
0o,
'"
WATTS/IN 2 1"C
S " FIN THICKNESS, INCHES
h :
SURFACE HEAT
7)
FIN EFFECTIVENESS
TRANSFER COEFFICENT
"
a : TURNING
d
'=
L~NE
f£-
:
EFFECTIVE DIAMETER
ROUND FIN
D-d
b=2
FIGURE 18.32
SQUARE FIN
~~~-L~4-L~~~7 ~
RATIO Old
0:0.564 E-d/2
o:
EFFECTIVE OIAMETER
OFFIN= 1126E
FIN EFFECTIVENESS NOMOGRAM FOR
FLAT, UNIFORM THICKMESS FIN
Fin effectiveness can be computed by means- of the nomogram
shown in Figure 18.32. The typical sequence of proceeding through
the nomogram is indicated by the encircled numbers adjacent to the
scales.
Example of Use of Fin Effectiveness Nomogram
Given:
Problem:
- Stack composed of 3" x 3" square painted aluminum
fins, each %4 inch thick.
- Effective stud hex diameter d = 0.59 inch.
- 1 inch spacing between fins
- 300 LFPM air velocity
- Fin temperature at stud = 100°C.
- Ambient air temperature = 40°C.
Determine total heat transfer of each fin.
537
SCR MANUAL
Solution:
1st Step: Determine total heat transfer coefficient.
transfer coefficient hr = .0024 w/in. 2 °C
Section 18.3.1.1.2.
Convection heat transfer coefficient he =
per example in Section 18.3.1.1.4.
Total heat transfer coefficient = hr + he
h = .0134 w/in. 2 DC.
Radiation heat
per example in
.011 w/in. 2 °C
= .0024 + .011,
2nd Step: Determine fin effectiveness factor from nomogram.
b= 0.564E - d/2 = (0.564) (3) - 0.:9 = 1.39"
D/d = 1.12~ X 3 = 5.64
0.9
For h = .0134 w/in. 2 °C and thickness 8 = .0155 inch.
ex = 0.58 as indicated by the dashed line on the nomogram. Through b = 1.39 and ex = 0.58, a line is extended
to the graph where D/d = 1. Projecting horizontally on
this graph to D/d = 5.64, 7J is found to be 0.67.
3rd Step: Determine total heat transfer.
Total heat transfer q = h A 7J ~T
q = (.0134 W/in.2 0c) (18 in. 2) (0.67) (100 - 40°C)
= 7.2 watts per fin.
For fin materials other than copper or aluminum, use the "copper"
scale on the nomogram by multiplying the actual fin thickness by the
ratio of the thermal conductivity of the material being considered to
the thermal conductivity of copper. Thus, for a 1Al inch steel fin, enter
axis 2 on the copper scale at 0.125 inch x 1.16/9.77 =0.015 inch.
Thermal conductivities of several commonly used fin materials are
given in Table 18.9.
>
Material
Aluminum
Brass
(70 Cu, 30 Zn)
Copper
Steel
Density
(lbs/in.3)
Heat Capacity (c)
(watt-sec.llb. 0c)
Thermal
Conductivity (k)
(watts/in. 0c)
0.098
0.30
407
179
5.23
2.70
0.32
0.28
175
204
9.77
1.16
TABLE 18.9 THERMAL PROPERTIES OF HEAT EXCHANGER MATERIALS
In general, it will be found that fin thickness should vary approximately as the square of the fin length in order to maintain constant fin
effectiveness. Also, a multi-finned assembly will generally have superior
fin effectivenss and will make better use of. material and weight than a
single Hat fin.
538
MOUNTING & COOLING THE POWER SEMICONDUCTOR
18.3.1.1.6 Typical Example of Complete Fin Design
Given:
Problem:
- Four C35 SCR's with %6" hex and %"-28 thread are
operated in a single-phase bridge at 10 amperes DC
maximum each. The specifications for this rectifier indicate that at this current level each SCR will develop 16
watts of heat losses at its junction and that for satisfactory service at this current level, the stud temperature should be maintained below 92°C. The maximum
ambient temperature is 40°C and free convection conditions apply.
Design a stack of fins to adequate cool the four SCR's
in this bridge circuit.
Solution:
1st Step: Determine maximum allowable fin temperature at radius
of stud hex. From Table 18.2, the thermal resistance from
stud to fin for a joint with lubricant is 0.25°C/watt maximum. The maximum fin temperature therefore must not
exceed 92°C - (0.25°C/watt x 16 watts) = 88°C.
2nd Step: Estimate required fin designs based on space available:
6" x 6" painted vertical fins at one inch spacing. Material
.08 inch thick steel. Assume all cell losses are dissipated
by fin.
3rd Step: Determine surface heat transfer coefficient and fin effectiveness of estimated fin design:
Radiation (from Nomogram in Figure 18.28)
TG = 88 ; 40 = 64°C
hr
= .00145 w/in.2 °C
Free Convection (from Nomogram in Figure 18.29)
h"
= .0037 w/in. 2 °C
htotal = .0052 w/in. 2 °C
Fin Effectiveness (from Nomogram in Figure 18.32)
D = 1.128E = 1.128 X 6 = 6.768
d = 0.57
.6.T = 88 - 40 = 48°C
b = 0.564E - d/2 = 0.564 X 6 - 0.:7 = 3.0
Did =
60~5~ = 11.8; 8 Cu =
(.08)
!:~~ =
.0095 in.
Using these parameters in the nomogram, 7J = 55%.
4th Step: Determine total heat transfer for estimated fin.
q=hA7J.6.T
= (.0052) (6 x 6 in. 2 ) (2 sides) (0.55) (86 - 40°C)
= 9.7 watts
5th Step: Determine error in approximation. Re-estimate fin requirements, and recalculate total heat transfer. In this example,
the capabilities of the initial fin design fell considerably
below the requirements of 16 watts. To sufficiently in-
539
SCR MANUAL
crease the heat transfer, a lJ4" thick copper fin would be
needed. Alternately a thinner fin of larger area could be
used.
From the above example it is seen that the fin heat transfer or
thermal resistance is a function of temperature as well as fin material,
size and geometrical configuration. Because of this fin operating temperature conditions must always be known regardless of procedure
used to select fin, i.e., from a manufacturer's catalog (see Tables 18.10,
18.12) or from the above design procedure. To give the designer some
perspective on the subject, Figure 18.32 is shown for the power tab
package at typical operating conditions. Similar curves can be made
up by using the above design procedure for thicker fins and other SCR
packages such as the power tab.
80
60
;.
~ 40
~,
.
I
\Te ·.5°1
20 , . - -
in
II!
.
TA=40·C, 1/2 SINE WAVE
\
\
lj
:!"
~:g~ ~~T~%~~~~e~~~fsEL
"i'85"e
\
..J
~
10
....
%
8
"
6
...
;;:
,Te-70 oe
1"1.1
I I t
FIGURE 18.33 HEAT EXCHANGER SQUARE FIN DESIGN FOR POWER TAB PACKAGE (SINGLE FIN)
18.3.1.2 Example of Calculating the Transient Thermal Impedance
Curve for aSpecific Heat Exchanger Design
Problem: A cell is mounted on a painted copper fin Vi6" thick and
4" on a side. The fin is subjected to free convection air conditions. Find
the transient thermal impedance curve for this fin. Assume that the
temperature throughout the heat exchanger is uniform even under
transient loading, thus permitting the heat exchanger to be represented
by a single time constant. This is a good assumption for fins of relatively thick cross-section and fin effectiveness close to unity. This
approach also assumes that the thermal capacity of the heat exchanger
is large compared to the thermal capacity of the cell.
Solution: From fin design curves
.005 watt/inch2 °C
he
.005 watt/inch2 °C
hr
h tota1
hr + he
.010 watt/inch2 °c
Fin thermal conductance k = h X A = .01 X 4 X 4 inches2
X 2 sides = 0.32 wattrC
=
=
=
Fin thermal resistance Of
540
=
= ~ = 0.~2 = 3.1 °C/watt
MOUNTING & COOLING THE POWER SEMICONDUCTOR
.
.
175 watt-seconds
FID thermal capacIty C = cpV =
lb 0C
X 0.32 Ib/in3
X
.
4.
4 ID.
X
ID. X
11·
716
ID.
= 56 watt-seconds
0C
Thermal RC time constant = 3.1 °C/watt X
56 watt-seconds
0C
= 174 secon1
High I
> 100
Stud/Pr~ss
Pac
Flat Base
.4- .04
a) Flat Fin
a) Extruded
Predominant
a) Flat Fin
Aluminum
Heat Exchanger b) Formed Flat b) Formed Flat
Fin
Type
Fin
(Convection
& Forced)
(Convection) c) Extruded
Aluminum
b) Liquid
(Convection
Cooling
& Forced)
TABLE 18.11
HEAT EXCHANGER TYPE GUIDE
18.3.2.1 Press Pak Vs Stud
Although not generally thought of as a heat exchanger question,
the decision of which package type to use has a direct bearing on
high current heat exchanger trade-offs. This relationship is clearly seen
by Figure 18.34, where heat exchanger volume requirements for free
convection press pac heat exchangers is 75% of stud heat exchangers.
The savings become more pronounced when comparing forced convection exchangers where the press pak sink has a 100% volume
advantage over its stud counterpart.
'.0
~
~
~
2.0
~
~
l..
l_ ."
1.0
"
"
.8
~
I
~
~ ::.....
I'-...
$~. ~ ~
~~~
~~
~.A.
NOTES:
I. EXTRUDED BLACK ANNDOZED
ALUMINIUM EXCHANGERS.
I-
1"....
~.. o""
~~
o"'o..
If'"'"
/)
~J..oe. ~
0>",..
""
j~
2. FORCED COOLED AT IPOO LF.M.
C'~~
3. TA =40"C.
4. PRESS PAK EXCHANGER DOUBLE
SIDE COOLED
~
20
t--t--
r--....
I
10
~ .....
"
40
60 80 100
HEAT EXCHANGER VOLUME -CUBIC INCHES
200
400
600
1,000
800
FIGURE 18.34 STEADY STATE THERMAL RESISTANCE Vs HEAT EXCHANGER VOLUME
542
MOUNTING & COOLING THE POWER SEMICONDUCTOR
Since heat exchanger volume has a direct relationship to heat
exchanger cost, the relationship should not be taken lightly. When
considering the total thermal circuit, the savings of press paks over
stud packages becomes even more significant due to further reductions
in thermal resistance within the semiconductor package of the press
pak over the equivalent silicon when mounted in a stud package, as
shown in Table 1B.12.
Silicon Sub-Assembly
Stud Rating
AmpsRMS
110
-
RaJO - °C/W~
Press Pak
(Double Side Cooled)
L;lbU
L;;jbU
-
- -- -----------
0.3
C180
---- 0.14
235
RaJO - °C/W
TABLE 18.12
Stud
0.135
C3BO
----------0.095
PRESS PAK Vs STUD PACKAGE THERMAL RESISTANCE
18.3.2.2 Free Vs Forced Air Convection
Forced cooling permits a four to one reduction in volume of heat
exchangers. If this was the only consideration all high current SCR's
would be forced cooled. Unfortunately the decision is not an easy one
when equipment reliability considerations are factored into the decision. Blowers and fans due to their mechanical nature reduce equipment reliability. To compensate for reduced reliability when using
blowers, designers have two options. Back-up or tandem blowers can
be used to provide for blower failures. In addition the improved SCR
cooling provided by blower operation may be used to provide for
lower SCR operating junction temperatures. Thus blower unreliability
is compensated by increased SCR reliability (see Chapter 19).
Blower requirements are determined from heat exchanger requirements. The trade-off in air flow vs thermal resistance for a typical heat
exchanger is shown in Figure 18.35. Note that the knee of the curve
1\
o.
• \
o.
•
o.
i
z
0
0.4
t
~
\,
I""
8
o.3
w
~
""'-
o. 2
EXCHANGER EXTRUDED AL
BLACK, ANODIZED
rAKEFIELD PP3849. 631N 3
~
" '"
100
200
400
.......
800
800 1000
AlIt fLOW-U.AIt FUT PEIlt IIIIINUTE.
FIGURE 18.35
PRESS PAK HEAT EXCHANGER THERMAL RESISTANCE Vs AIR FLOW
543
SCR MANUAL
beyond which diminishing returns commence to take place occurs at
500 LFPM. Blower selection procedure is outlined by use of curves
simimr to those shown in Figure 18.36. Blower head vs air How curves
can be likened to load lines, correspondingly the exchanger head vs
air How curve is drawn similar to a transistor or diode characteristic,
the intersection of the two curves determine the operating point.
°O~~~--~40~~&---~--~--~~~'~
BLOWER a EXCHANGER AIR fLOW - CUBIC FEET PER IIINl!TE
FIGURE 18.36 DETERMINING HEAT EXCHANGER·BLOWER AIR FLOW OPERATING POINT
For example the larger blower "A" intersects the heat exchanger
curve at 82 CFPM thus resulting in a low thermal resistance at the
operating point at 1 and l' of 0.14°C/watt. Since blower A provides
an operating point well below the knee of the RasA vs CPFM curve,
little is to be gained by going to a larger fan. While the smaller fan's
operating point is at the ResA knee it may be a useful operating point
since it provides better than a 2: 1 improvement over the free convection RasA value. Both the blower and heat exchanger curve data is
provided on manufacturer's data sheets with the units shown such
that a direct matching of blower to exchanger requirements is easily
accomplished.
Care should be taken regarding the following additional factors
when selecting exchangers and blowers.
1. Blower noise
2. Altitude effects
3. Additional head losses due to. filters and ducting
4. Temperature rise in serial air How arrangements
t:. T
= [1.76 X PdiSSiPated]
CFPM
Air How measurement techniques as well as additional considerations are discussed in det:ail in References 7, 8, 9, 10 and 11. For
detailed manufacturer's literature consult the firms listed in Tables
18.12 and 18.13.
544
MOUNTING & COOLING THE POWER SEMICONDUCTOR
1. Astro Dynamic Inc.
Second Ave. NW Industrial
Park
Burlington, Mass. 01803
2. Astrodyne Inc.
207 Cambridge Street
Burlington, Mass. 01803
3. International Electronic
Researcb Corp. *
135 W. Magnolia Blvd.
Burbank, California 91502
4. George Risk Industries Inc.
67215th Avenue
Columbus, Nebraska 68601
5. Thermaloy Co. *
8719 Diplomacy Row
Dallas, Texas 75247
6. Tor Inc.
P. O. Box 8
Irwindale, California 91706
7. Vemaline
Wyckoff, New Jersey
8. Wakefield Engr. Inc.*
Wakefield,Mass.01880
9. Seifert Electronic
Ing. Rolf Seifert
5830 Schwelm
Postfach 270
West Germany
10. Hans Schaffner
Elektronisches Bauteile
4708 Luterbach
Switzerland
*Supplies of liquid cooled heat exchangers in addition to extruded
designs.
TABLE 18.12
MANUFACTURERS OF EXTRUDED HEAT EXCHANGERS
1. IMC Magnetics Corp.
Eastern Division
507 Main Street
Westbury, New York 11591
2. Pamotor, Inc.
770 Airport Blvd.
Burlingame, California 94010
3. Rotating Components
1560 5th Avenue
Bay Shore, New York 11706
TABLE 18.13
4. Rotron Inc.
Woodstock, New York 12498
5. The Torrington Mfg. Co.
Torrington, Conn.
6. W. W. Grainger Inc.
3812 Pennsylvania Avenue
Pittsburgh, Penn. 15201
7. Parker Hannifin UK (Ltd.)
Tube & Hose Fittings Division
Haydock Pk. Rd.
Derby, England
MANUFACTURERS OF BLOWERS
545
SCR MANUAL
18.3.2.2 Liquid Cooling
Liquid cooling is gaining increasing acceptance in very high
power applications because it offers the lowest values of RasA in combination with extremely small volume requirements when using tap
water for the liquid coolant. It is a natural progression from forced
convection air cooling to forced liquid cooling. Space limitations do
not allow adequate treatment of all facets of the subject. This section
will concentrate on two aspects of the subject while prOviding references to further aid the designer in remaining areas,
18.3.2.2.1 Heat Exchanger Selection
Many of the same variables that determine air cooled. exchanger
selection also hold for liquid cooled exchanger selection criteria. The
major variables are listed below.
a) Thermal resistance - RacA vs liquid How rate
b) Pressure drop vs liquid How rate
c) Exchanger material as it relates to atmospheric corrosion
d) Size, weight and cost
e) Ease of assembly and Hexibility
Additional factors not generally met with air cooling consist of:
a) Susceptability to plugging (size of liquid passages)
b) Material compatability to liquid
c) Method of liquid conenctions, i.e., hose clamps, pipe fittings,
etc.
These latter factors can be critical to proper exchanger selection,
i.e., liquid passage ways should be a minimum of 3fs" ID; unplated
copper and aluminum is not recommended with water. Liquid connections, while not critical, should allow for tight leak free connections
not generally a problem with water but definitely a consideration if
liquid having low surface tension are contemplated and/or systems
with high liquid pressures.
General Electric has two liquid heat exchanger designs employing two different concepts as shown in Figure 18.37(a) and (b).
546
MOUNTING & COOLING THE POWER SEMICONDUCTOR
CROSS SECTION A-A
POST
LIQUID
PASSAGEWAYS
ANODE POST
(8) G6 Liquid Cooled Heat Exchanger
(b) G5 LiqUid Cooled Heat Exchanger
FIGURE 18.37
HIGH PERFORMANCE LlQUIO COOLED HEAT EXCHANGERS
The G6 liquid cooled exchanger or post as it is sometimes called,
provides for both turbulent How and large heat transfer area. This
combined with a relatively large, short liquid passage way provides
low values of ResA and pressure drop. Furthermore, it transfers heat
547
- SCR MANUAL
to both post ends thus allowing most compact doubler and series
arrangement of SCR's, with a minimum of liquid connections.
The C6 heat exchanger is available in conjunction with any of
the Series C3_ SCR's in many different assembly configurations, i.e.,
AC switches, doublers, diode SCR combinations, etc.
The C5 heat exchanger employs a unique recessed liquid passage
way as seen in Figure IB.37(b). The short Vee passage way combined
with the relatively large diameter provides low pressure drops. While
having a low pressure drop characteristic the Vee design also greatly
reduces the SCR package thermal resistance due to the proximity of
the Vee to the silicon sub-assembly. This, together with the enhanced
heat transfer properties that take place at the bottom of the Vee due
to liquid turbulence, enables the C5 heat exchanger to have an
extremely low overall thermal resistance liquid to junction. The C5
heat exchanger is available in conjunction with any of the Series C5_
and CB_ SCR's. Like the C6, the C5 is available in a broad range
of packaged assemblies.
Table IB.12 lists suppliers of liquid cooled heat exchangers.
18.3.2.2.2 LiquillSelection and Requirements
Pure, deionized, water is the best heat transfer medium when
maintenance of closed loop systems are compatible with overall system
design objectives and the water can be maintained at above freezing
temperature. Many industrial systems employ raw tap water due to
the lower initial cost.
The quality of raw water for cooling systems, where heat exchangers are not employed, should have the following purity:
1) A neutral or slightly alkaline reaction, i.e., a pH between 7.0
and 9.0.
2) A chloride content of not more than 20 parts per million: a
nitrate content of not more than 10 parts per million; a sulphate content of not more than 100 parts per million.
3) A total solids content of not more than 250 parts per million.
4) A total hardness, as calcium carbonate, of not more than 250
parts per million.
Chemical analyses are not always available to assist in an appraisal
of cooling water. In such cases, an electrical resistivity measurement
of the water will provide a satisfactory guide to the total amounts of
dissolved solids. Water having a resistivity of 2,500 ohm-centimeters
or higher, when measured at about 25°C, is usually satisfactory as a
coolant. The approximate amount of total dissolved solids can be determined by the equation:
Total Dissolved Solids in Parts per Million
=
MO,OOO
Specific Resistivity in Ohm-Centimeters
Raw water insulating hose connections to the converter or
ungrounded heat exchangers should be long enough to reduce the
leakage current to a tolerable level (IB" or greater), or electrolytic
targets, should be used at the hose fittings.
548
MOUNTING & COOLING THE POWER SEMICONDUCTOR
Reference 6 provides information on electrolytic targets, leakage
currents and rust inhibitors for closed loop systems.
Whenever the coolant temperature is below ambient cabinet temperature, the possibility of accelerated external corrosion and electrolysis must be considered due to condensation of water vapor in the
ambient air taking place on the semiconductor insulating surfaces.
Dehumidifiers or water tempering are possible solutions to the problem.
Water tempering raises the water temperature slightly by mixing
hot water with the cold supply water such that the coolant temperature
is above air ambient. The dehumidifier removes moisture from the air
thus substantially lowering the dew point.
When employing antifreeze solutions or oil coolants in closed loop
systems, heat exchanger performance is degraded due to the inferior
properties of cooling fluids other than water. Reference 13 provides
formula for calculating the performance characteristic of the GE G6
heat exchanger with liquids other than water as a function of the liquid
properties, i.e., viscosity, specific heat and density.
Table 18.13 contains an abbreviated list of tubing and tube fitting
manufacturers.
Imperial Eastman
6300 W. Howard Street
Chicago, Illinois 60648
Norton Plastics & Synthetics Div.*
Akron, Ohio 44309
Parker & Hannifin
300 Parker Drive
Ostego, Michigan 49078
*Tubing manufacturer only
TABLE 18.13
TUBING & TUBE FlnlNGS MANUFACTURERS
18.4 MEASUREMENT OF CASE TEMPERATURE
Heat exchanger design should be checked in the prototype equipment. A 10 or 12 mil thermocouple wire should be used. A copperconstantan thermocouple junction is suggested. The thermocouple junction should be carefully attached to the SCR case as indicated iIi.
Figure 18.38.
549
SCR MANUAL
THERMOCOUPLE
LEADS
\
HEAT
EXCHANGER
(a) STUD
SPRING CLAMPS
(b) FLAT BASE
THERMOCOUPLE LEADS
THERMOCOUPLE
COOLING FIN
~
(e) PRESS FIT
THERMOCOUPLE
LEADS
' - - - = - - - i + - - C I . A M P INSULATOR
EXCHANGER
(d)
FIGURE 18.38
PRESS PAK
PREFERRED LOCATION FOR MOUNTING THERMOCOUPLE FOR CASE.
TEMPERATURE MEASUREMENTS
While soldering may be used for the method of attachment to
low current devices, it is not practical or recommended for high current
devices.
Use of an Amalgam provides the advantages of the soldering
method without its danger. Its major drawback is the additional
preparation time. It is far superior to fastening by peening.
18.4.1 Materials Used
Gallium - 32%; tin - 18%; copper - 50%.
325 mesh copper and tin are available from A. D. Mackay Inc.,
198 Broadway, New York, New York 10038. Small balls of gallium are
available from Eagle-Picher Industries, Inc., Quapaw, Oklahoma
74363.
550
MOUNTING & COOLING THE POWER SEMICONDUCTOR
18.4.2 Preparation
A small teHon bowl and teHon IDlxmg stick is used to prepare
Amalgam. A small amount of gallium is shaved from a gallium ball
and copper is added, not quite twice as much as gallium, and a small
amount of tin is added, keeping in mind the proportions specified above,
but no attempt is made at actual measurement of these quantities.
The gallium at normal room temperature is solid but will liquefy
at slightly higher temperatures such as from body temperature as when
held between fingers. Pressure also tends to liquefy the gallium as
when being forced into holes to hold thermocouple. In the type of action
sought for holding thermocouples if too much gallium is used the mix
will be excessively shiny and more copper should be added to cause
the gallium to move from the more liquefied state to a more paste
consistency.
Too little tin makes the mix dark. Tin acts as an inhibitor which
retards initial set and thus facilitates packing.
The copper is used to give the mix body.
The setting time of mix is about one hour. The hardening time is
approximately twenty-four hours.
The temperature at the points shown in Figure 18.38 closely
approximates the temperature of the case immediately below the junction, which is usually inaccessible once the device is mounted to 'an
exchanger. The point of measurement on the case should be shielded
from any forced air which might cause localized cooling, and the leads
should be kept out of any How of cooling air since they can provide a
heat How path which will lower the temperature at the thermocouple
junction.
Use of the foregoing procedures to produce a well-engineered
cooling exchanger design for SCR's can pay big dividends in reliable operation, low material costs, and minimum space and weight
requirements.
Unless carefully calibrated leads and instruments are available, a
thermocouple bridge rather than a pyrometer should be employed.
Care should be taken to keep the thermocouple leads out of electric
fields that might induce error voltages in the leads.
As an alternative to using a thermocouple, temperature indicating
waxes and paints bearing such trademarks as "Thermocolors" (manufactured by Curtiss-Wright Research Corporation) and "Tempilaq"
(manufactured by Tempi! Corporation, New York City) can be used to
indicate whether the case exceeds a specific level of temperature.
Careful attention should be given to the manufacturer's instructions
for using this type of temperature indicator to prevent mis-application
and errors. Temperature indicating paints and waxes are particularly
useful in high electrical fields where substantial errors may occur in
electrical measurement techniques or where the exchanger is inaccessible for thermocouple leads during the test, such as on the rotor of a
rotating machine. Care must be taken in applying paints so their
presence does not materially affect the emissivity of the surface.
551
SCR MANUAL
REFERENCES
1. "Heat Transfer," Vol. II, Jakob, John Wiley & Sons, Inc., New
York, N. Y., 1957.
2. "Principles of Engineering Heat Transfer," Giedt, VonNostrand
Co., New York, N. Y., 1957, p. 218.
.
3. "Heat Transmission," McAdams, McGraw-Hill Book Company,
New York, N. Y., 1942:
4. "Thermal Mounting Considerations for Plastic Power Semiconductor Packages;" R. E. Locher, Application Note 200.55,* General Electric Company, Auburn, N. Y.
5. "A Variety of Mounting Techniques for Press-Fit SCR's and Rectiners," J. C. Hey, Application Note 200.32, * General Electric Company, Auburn, N. Y.
6. "Mounting Press Pak Semiconductors," B. W. Jalbert, Application
Note 200.50, General Electric Company, Auburn, N. Y.
7. "How to Select Fans," James W. Fry, Product Design & Value
Engr., Sept./Oct. 1964. Available from Torrington Mfg. Co.
8. "Air Circuit Design," Kenneth A. Merz, Machine Design, June 28,
1956. Available from Torrington Mfg. Co.
9. "Noise in Air-Moving Systems," Roberty J. Kenny, Machine Design,
September 26, 1968. Available from Torrington Mfg. Co.
10. "Design For Quiet," Machine Design, September 14, 1967. Available from Torrington Mfg. Co. An excellent comprehensive overview of all facets of noise and its measurement in both motors and
blowers.
11. "Air Flow Measurement," Erich J. O. Brandt, Product Engineering,
June 1957. Available from Torrington Mfg. Co.
12. Handbook of Heat Transfer Media, Paul L. Geiringer, Reinhold
Publishing, New York.
13. "Liquid Cooling of Power Thyristors," F. B. Golden, IEEE 1970
IGA Conference Record.
*Refer to Chapter 23 for availability and ordering information.
552
SCR RELIABILITY
19
SCR RELIABILITY
19.1 INTRODUCTION
Reliability is not new as a concept, but the language and techniques relating to its treatment have continued to develop as technology
has advanced and become increasingly complex. The need to define
reliability as a product characteristic has expanded as the newer technologies have moved from laboratory to space to industry to home.
The steel mill calculates the cost of down time in thousands of dollars
per minute; the utility is sensitive to the low tolerance level of its
customers to interruptions in service; the manufacturer of consumer
equipment relies on a low incidence of in-warranty failures to maintain
profitability and reputation.
The complexity of equipment, on the one hand, and the development of new components on the other, have forced industry to invest
considerable effort in finding means for controlling and predicting
reliability. The efforts, in many cases, were accelerated by the desire
of the military to evaluate, and improve where necessary, the reliability
of new devices, which offered the promise of improvements in size,
weight, performance, and reliability in aerospace and weapons systems.
One new device so favored was the SCR, the first of the thyristor
family of devices to be commercially available.
The first SCR, the General Electric C35, was successfully qualified
to the first SCR military specification only two years after it was made
commercially available. At approximately that same time, a specification was finalized to which this same device was qualified as part of
the highly publicized Minuteman missile high reliability program.
These programs, and others that followed, contributed a great deal to
the knowledge and understanding of the inherent reliability of semiconductors such as the SCR, and of those factors in design, rating,
process control and application that effectively determine the reliability
achieved. As a result the General Electric Co. has been able to develop
and produce a wide variety of reliable thyristor devices tailored to the
particular needs of various fields of application.
19.2 WHAT IS RELIABILITY?
Reliability may be defined as the probability of performing a
specific function under given conditions for a specific period of,time.
Reliability is a measure of time performance as opposed to quality,
which is a measure of conformance to specified standards at a given
point in time. Although system reliability is influenced by factors such
as the selection and design of circuits. Discussion in this chapter is limited to the effects of component part reliability. In addition, the assumption is made that the parts are properly applied, and that they are not
subject to stresses that exceed rated capability;
553
SCR MANUAL
19.3 MEASUREMENT OF RELIABILITY
In the case of large systems, the common unit of reliability measurement is MTBF, or Mean Time Between Failures. MTBF expresses
the average time in hours that the system operates between failures,
providing a basis for estimating the cost of system maintenance. The
MTBF measurement further contributes in the establishment of preventive maintenance scheduling and in estimating productivity as a
function of availability, which is that percentage of time that the system can be expected to be productively operable.
The reliability of a system is based on the summation of the reliabilities of all the parts that make up the system. This process is made
complex by factors such as the need for weighting based on the effect
on total system performance of the failure of a particular component or
circuit, and the assignment of correction factors to compensate for stress
levels applied. If these complexities are ignored, and if a further simplification is made in the form of the assumption that the failure rates of
the components are constant over time, then Failure Rate is the reciprocal of MTBF, and system MTBF is the reciprocal of the sum of the
Failure Rates of the component parts.
19.3.1 Failure Rate
An individual component part, such as a semiconductor, does not
lend itself to reliability measurement in the same manner as does a
system. For this reason, the statistical approach to estimating device
reliability is to relate the observed performance of a sample quantity
of devices to the probable performance of an infinite quantity of similar
devices operated under the same conditions for a like period of time.
The statistical measurement is based on unit hours of operation, using
a sampling procedure whose derivation takes into account the resolution with which the sample represents the population from which it
was drawn and the general pattern of behaviour of the devices observed
with time.
The sampling plan most commonly applied to semiconductors is
given as Table C-l of Mil-S-19500E, and is shown here as Figure 19.1.
"Failure Rate" is a commonly used term, generally applied interchangeably with LTPD (Lot Tolerance Percent Defective), which is also
called "Lambda" when used in connection with a one thousand hour
test period. The table given permits calculation of failure rate at the
_"ninety percent confidence level as a function of the number of devices
, observed and the number of failures occurring.
According to the sampling _plan, satisfactory operation of 231
devices for 1000 hours is indicative that the failure rate is no greater
than 1.0% per 1000 hours at a 90% confidence level. If it were desired
to demonstrate a maximum failure rate of 0.1 % per 1000 hours, the
minimum sample would be 2,303 devices with no failures allowed. This
could also be demonstrated with a sample of 3,891 samples, allowing
one failure. In either case, a successful test would be the equivalent
of demonstrating an MTBF of 1,000,000 hours for a system made up
solely of the devices under test. Several points become evident in these
observations:
554
SCR RELIABILITY
(a) It would be extremely difficult to perform an accurate test
demonstration to verify failure rate even 1.0% since the test equipment and instrumentation must have a still greater MTBF in order not
to adversely affect the test results. The problem compounds as the
failure rate being tested for is lowered. Not only is test equipment
complexity increased, but its MTBF must be increased at the same
time!
(b) The terminology "Failure Rate" is perhaps a poor choice of
words. To the reliability engineer it relates the performance of a limited number of observations to the probable performance of an infinite
population. To those not familiar with the statistics used, it unfortunately conveys the impression of actual percent defective.
19.4 SCR FAILURE RATES
Graphical presentations such as described· in Section 19.7 have
been found very useful to electronic device users as a guide for reliability predictions.
As an example of SCR reliability, a sample of approximately 950
pieces C35 type SCR's were subjected to full load, intermittent operation of 1000 hours duration in formal lot acceptance testing to MIL-S19500/108 from 1962 through 1970. Of these only one device was
observed to be a failure to the specification end point limits. The calculation of failure rate based on these results indicates the failure rate to
be no more than 0.41 % for 1000 hours at 90% confidence.
19.5 DESIGNING SCR'S FOR RELIABILITY
The design of reliable devices is concerned with the assurance
that performance related characteristics remain within specified tolerances over the useful life of the devices. This relates particularly to
thermal and mechanical design.
In the case of thermal design, the stability of thermal transfer
characteristics are important for the reason that junction temperature
is the major application limitation. The deterioration of the thermal
path can lead to thermal runaway and device destruction. Interface
materials scientifically selected for matched coefficients of expansion
compatible with the rated range of temperatures are necessary to
reduce the likelihood of metal fatigue.
Mechanical reliability requires the use of rigid assemblies of low
mass, low moments of inertia, and the elimination of mechanical resonances in the normal ranges of vibration and shock excitation. Equally
critical is the design of the protection of the junction surface, whether
it be hermetic seal or passivation. Since degradation failures are mainly
manifestations of changes at the junction surface, reliability is closely
related to the integrity of the surface protection.
Lower costs and volume processing can be accomplished through
new techniques without compromise of reliability. This has been exemplified by the CI06 solid encapsulated SCR. Effective silicon dioxide
passivation and the development of a compatible encapsulant have
eliminated the need for glass to metal hermetic seals. Life test results
indicate that the CI06 is as reliable as hermetically sealed devices.
Other examples are the C122, SC141 and SC146 which use glassivation as an effective means of passivation.
555
~
a>
Minimum size of sample to be tested to assure, with a 90 percent confidence, a Lot Tolerance Percent Defective or X no greater than the LTPD specified. The minimum
quality (approximate AQL) required to accept (on the average) 19 of 20 lots is shown in parentheses for information only.
Maximum
Percent
Defective
(LTPD) or
X (1)
20
15
10
7
5
3
2
1.5
1
0.7
D.5
0.3
0.2
0.1
Rejection
Number
Acceptance
Number
1
0
11
(0.46)
15
(0.34)
22
(0.23)
32
(0.16)
45
(0.11)
76
(0.07)
116
(0.04)
153
(0.03)
231
(0.02)
328
(0.02)
461
(0.01)
770
(0.007)
1152
(0.005)
2303
(0.002)
2
1
18
(2.0)
25
(1.4)
38
(.94)
55
(.65)
77
(.46)
129
(.28)
195
(.18)
258
(.14)
390
(.09)
555
(.06)
778
(.045)
1298
(.027)
1946
(.018)
3891
(.009)
3
2
25
(3.4)
34
(2.24)
52
(1.6)
75
(Ll)
105
(.78)
176
(.47)
266
(.31)
354
(.23)
533
(.15)
759
(.11)
1065
(:080)
1777
(.046)
2662
(.031)
5323
(.Ql5)
4
3
32
(4.4)
43
(3.2)
65
(2.1)
94
(1.5)
132
(1.0)
221
(.62)
333
(.41)
444
(.31)
668
(.20)
953
(.14)
1337
(.10)
2228
(.061)
3341
(.041)
6681
(.018)
5
4
38
(5.3)
52
(3.9)
78
(2.6)
113
(1.8)
158
(1.3)
265
(.75)
398
(.50)
531
(.37)
798
(.25)
1140
(.17)
1599
(.12)
2667
(.074)
3997
(.049)
7994
(.025)
6
5
45
(6.0)
60
(4.4)
91
(2.9)
131
(2.0)
184
(1.4)
308
(.85)
462
(.57)
617
(.42)
927
(.28)
1323
(.20)
1855
(.14)
3099
(.084)
4638
(.056)
9275
(.028)
7
6
51
(6.6)
68
(4.9)
104
(3.2)
149
(2.2)
209
(1.6)
349
(.94)
528
(.62)
700
(.47)
1054
(.31)
1503
(.22)
2107
(.155)
3515
(.093)
5267
(.062)
10533
(.031)
8
7
57
(7.2)
77
(5.3)
116
(3.5)
166
(2.4)
234
(1.7)
390
(1.0)
589
(.67)
783
(.51)
1178
(.34)
1680
(.24)
2355
(.17)
3931
(.101)
5886
(.067)
11771
(.034)
9
8
63
(7.7)
85
(5.6)
128
(3.7)
184
(2.6)
258
(1.8)
431
(1.1)
(.72)
648
864
(.54)
1300
(.36)
1854
(.25)
2599
(.18)
4334
(.108)
6498
(.072)
12995
(.036)
10
9
69
(8.1)
93
(6.0)
140
(3.9)
201
(2.7)
282
(1.9)
471
(1.2)
709
(.77)
945
(.58)
1421
(.38)
2027
(.27)
2842
(.19)
4739
(.114)
7103
(.077)
14206
(.038)
11
10
75
(8.4)
100
(6.3)
152
(4.1)
(2~ . (2~ ~.2L
218
306
511
770
(:80)
1025
(.60)
1541
(.40)
2199
(.28)
3082
(.20)
5147
(.120)
7704
(.08)
15407
Minimum Sample Sizes
. -
(1) The life test failure rate lambda (X) shall be defined as the LTPD per 1000 hours.
FIGURE 19.1
LAMBDA SAMPLING PLAN AT 90% CONFIDENCE
(.O~
~
::0
!
~
SCR RELIABILITY
19.6 FAILURE MECHANISMS
Failure mechanisms are those chemical and physical processes
which result in eventual device failure. The kinds of mechanisms that
have been observed in the semiconductor classification of component
devices are shown in the table of Figure 19.2. Also shown in the table
are those kinds of stresses to which each mechanism is likely to respond.
If more than a few such failure mechanisms are, to any significant
degree prevalent in a given device type from a given process, it would
not be reasonable to expect to achieve the degrees of reliability that
have been demonstrated by many semiconductors. The dominant
mechanisms to which a device type may be susceptible will vary
according to the peculiarities of the design and fabrication process of
that device.
.. ..
M::CH, .NICU TEMPERATURE
w
~ 0:,,0 !.!
2
0
0"
!c
0
3 !c....en %en0'" ~::;
e~ % 0:OJ
Lm
~
• • • • •
ow
~~
FAILURE
MECHANI~M
en
STRUCTURAL FLAWS
-WEAK PARTS
-WEAK CONNECTIONS
- LOOSE PARTICLES
-THERMAL FATIGUE
ENCAPSULATION FLAWS
INTERNAL CONTAMINANTS
-ENTRAPPED FOREIGN GASES
-OUTGASSING
-ENTRAPPED IONIZABLE
CONTAMINANTS
- BASE MINORITY
CARRIER TRAPPING
-IONIC CONDUCTION
-CORROSION
MATERIAL ELECTRICAL FLAWS
-JUNCTION IMPERFECTION
METAL O\FFUSION
SUSCEPTABILITY TO
RADIATION
:;
0.o:!:
0
ELECTRICAL
IMISCELLANEOUS
,
....z
2
0
....>- ~
0: 6
'2
......
Wz
0:1~
Ci
1!
~Q iii
'"0:0:
...J
w
g
"
0
~~e
OZ
0.8
• •
•
•
• • • • • • •
• •
•
• •
•
•
• •
•
•
•
•
•
•
•
•
•
• • • •
• • •
•
• • •
•
•
FIGURE 19.2
~~
Q.!,!
•
•
•
•
Oen
00
..
0:
OJ
~
i
":r :
• • •
•
FAILURE MECHANISMS AND ASSOCIATED STRESSES
19.6.1 Structural Flaws
Structural Haws are generally considered to be the result of weak
parts, discrepancies in fabrication, or inadequate mechanical design.
Various in-process tests pedormed on the device, such as forward volt·
age drop at high current density levels and thermal resistance measurement, provide effective means for the monitOring of controls against
such Haws. These tests also provide a means for the elimination of the
occasional possible discrepant device.
The modes of failure generally associated with the mechanical
Haw category of failure mechanism for an SCR are excessive on-voltage
drop, failure to turn on when properly triggered, and open circuit
between the anode and cathode terminals. Because these types of
failure mechanism are relatively rare, the incidence of these modes of
failure is low.
557
SCR MANUAL
19.6.2 Encapsulation Flaws
Encapsulation flaws are deficiencies in the hermetic seal or passivation that will allow undesirable atmospheric impurities to reach the
semiconductor element. Foreign atmospheres, such as oxygen and
moisture, can react in such a way as to permanently alter the surface
characteristics of the silicon metal.
A change in surface conductivity is evidenced by gradual increase
of the forward and reverse blocking current. Because the SCR is a
current actuated device, it will lose its capacity to block rated voltage
if blocking current degrades beyond some critical point. This type of
mechanism may eventually result in catastrophic failure. The rate
of degradation is dependent mostly on the size of the flaw and the
level of stress, particularly temperature, that is applied.
In the case of hermetically sealed devices a sequence of fine and
gross leak testing can eliminate the occasional discrepant device. The
use of radiflo and bubble testing has been found very effective for
this means. Since the new plastic devices are solid encapsulated and
have no internal cavity, conventional methods of leak testing obviously
are no longer applicable; it has been necessary to develop new methods.
One of these methods is the "pressure cooker" type test (29PSIA, 121°C)
which has been found to be very effective in detecting devices with
defective passivation.
19.6.3 Internal Contaminants
The inclusion of a source of ionizable material inside a hermetically
sealed package, or under a passivation layer, can result in failure
mechanisms. These mechanisms are similar to those resulting from
encapsulation flaws if the inclusion is gross. If the inclusion is small,
as compared to the junction area, the amount of electrical change that
occurs is limited. Thus the increase in blocking current is not sufficient
to effect the blocking capacity of the device.
The mechanism need not be a permanent change in the surface
_ characteristics of the silicon. The apparent surface conductivity of the
silicon can be altered by build-up and movement of electrical charges
carried by the inclusions. The condition is often reversible, with recovery accomplished through the removal of electrical bias and the introduction of elevated temperature.
Because the SCR is a bistable, rather than a linear device, concern
for this category of failure mechanism arises oolyif the forward blocking current can increase to the point where forward blocking capability
is impared. The probability of occurrence is extremely low except for
the possible case of the small junction area, highly sensitive devices.
Even here, the mechanism is often negated through negative gate or
resistor biasing in the circuit.
Removal of devices containing undesirable internal contaminants
can effectively be accomplished by means of a blocking voltage burn-in
screen. Ionization of the contaminants under these conditions takes
place rapidly thus permitting a relatively short term burn-in. Detection
of the discrepant devices is accomplished by both tight end~point limits
and means to detect turn-on during the screen.
558
SCR RELIABILITY
19.6.4 Material Electrical Flaws
This category of failure mechanism involves, basically, imperfections in junction formation. Discrepancies of this nature are not generally' experienced with SCR's because of their relatively thick base
widths and because the blocking junctions are formed by the diffusion
process, which allows consistent control of both depth and uniformity
of junction. Initial electrical classification would effectively remove any
such discrepant devices.
19.6.5 Metal Diffusion
Of the possible failure mechanisms observed in semiconductors,
metal diffusion is the least significant. Though diffusion will occur over
a long period of time when two metals are in intimate contact at very
high temperatures, the rate at which it progresses is too slow to have
tangible effects during the useful life of the device or the system in
which it is applied. For example, many SCR's are gold diffused at
temperatures exceeding 800°C for times approaching two hours. In
this fashion it is possible to obtain desired "speed" characteristics. To
accomplish the equivalent gold diffusion at 150°C would require
approximately 3 X lOs hours (34,000 years).
19.6.6 Nuclear Radiation
Early studies, concerning the radiation tolerance of semiconductor
devices, indicated that thyristors were more susceptible to a degradation in electrical characteristics than bipolar transistors. This has more
recently been shown to be an invalid conclusion. To a great extent the
radiation resistance is determined by the Nand P base widths in a
thyristor and the base and collector widths in a transistor. The narrower the width, the less susceptible a device is to radiation. It is these
same widths however that determine the devices' blocking voltage
capability. Therefore, one should expect greater radiation resistance
from lower voltage devices than from high voltage devices, regardless
of whether they are thyristors or bipolar transistors. It must be kept
in mind that when selecting radiation resistant devices, it is the
designed blocking voltage that is critical, not the actual blocking voltage of the particular device.
The only true means for determining the actual tolerance of any
device to the effects of nuclear radiation is through actual radiation
exposure testing of that device. Approximate levels of SCR tolerance,
however, have been determined through various tests performed on
the General Electric C35 (2N685 series). Critical levels have been
shown to be 1014 nvt for fast neutron bombardment and 5 x 105 R/sec
for gamma radiation.
Fast neutron bombardment of the silicon results in permanent
damage to the crystal lattice, reducing minority carrier lifetime. Significant effects that appear between 1013 nvt and 1014 nvt are increased
gate current to trigger and, to a lesser degree, increased holding current, on voltage, and forward breakdown voltage.
559
SCR MANUAL
Although gamma radiation may also produce permanent ellects
on the SCR, it is expected that failure in the typical radiation environment would result first from fast neutron bombardment. Gamma radiation, however, produces high energy electrons by photoelectric and
Compton processes which create a leakage current during irradiation.
High pulse levels of irradiation can have the transient ellect of triggering the SCR on. At 106 R/sec, there is a fifty percent chance that the
General Electric C35 SCR will be triggered on.
19.7 EFFECTS OF DERATING
From the above, the most probable failure mechanism is degradation of the blocking capability as a result of either encapsulation Haw
(or damage) or internal contaminants. The process can be either chemical or electrochemical, and therefore variable in rate according to the
degree of temperature and/or electrical stress applied.
Thus it is possible by means of derating (using the device at stress
levels less than the maximum ratings of the device) to retard the process
by which the failure of the occasional defective device results. This
slowdown of the degradation process results in lower failure rate and
increased MTBF. Suppose, for example, that a sample of 778 devices
is tested under maximum rated conditions for 1000 hours with one
failure observed. The calculated Lambda (.\.) is 0.5 (see Table 19.1)
and the MTBF is 200,000 hours. H the failed device would have
remained within limits at the 1000 hour point because of lower applied
stresses, the calculated Lambda becomes 0.3 and the MTBF increased
to 333,000 hours.
The relationship of applied stress to General Electric SCR device
failure rate is shown graphically in Figures 19.3, 19.4 and 19.5. The
model that describes the relationship of these stresses as they relate to
failure rate, Lambda (A), is the Arrhenius Model. The Arrhenius Model
is given by:
Failure Rate, .\. = e A + B/Tj
where
A Failure rate expressed in % Per 1000 hours
T j = Junction temperature in degrees Kelvin
A and B Constants
The Arrhenius Model relationship has been successfully applied by the
General Electric Company to extensive life test data involving thousands of devices and millions of test hours. The data was obtained from
product design evaluations, military lot acceptance testing, and several
large scale reliability contracts.
A thorough examination of the data on all General Electric silicon
controlled rectifiers revealed that these three graphical presentations
could describe the results of derating on failure rate for the entire
family of SCR's with reasonable accuracy.
The use of these graphical presentations is quite straightforward.
Suppose for example, that it is desired to obtain the estimated failure
rate of a C35D under stress conditions of 200 volts peak and a junction temperature of 75°C. The circuit this device will be used in will
become inoperative when the electrical characteristics of the SCR
change to values outside of the specification limits. This exemplifies a
degradation definition of failure and signifies that the solid lines on the
=
=
560
SCR RELIABILITY
graphical presentations must be used. Since the rated junction temperature of the C35D is 125°C, Figure 19.4 must be used. Projecting
a horizontal line from the intersection of the 75°C junction temperature
ordinate and the applicable per cent of rated voltage surve (50% in
this example) we obtain an estimated failure rate of .08% per 1000
hours at 90% confidence. If, due to a change in the design of the equipment, only devices which failed catastrophically (opens or shorts)
would cause the equipment to become inoperable, the dashed curves
could be used. This would result in an estimated failure rate of .008%
per 1000 hours at 90% confidence.
0.001
0.0001
KEY
DEGRADATION DEFINITION OF FAILURE
- - - CATASTROPHIC DEFINITION OF FAILURE
NOTES:
I. FOR ESTIMATING FAILURE RATES OF DEVICES
SUBJECTED TO STRESS SCREENING TESTS,
DIVIDE THE FAILURE RATE OBTAINED FROM
THIS CURVE BY 10.
2.THIS CURVE IS TO BE USED FOR ESTIMATING
• FAILURE RATES OF SILICON CONTROLLED
RECTIFIERS WITH A MAXIMUM RATED JUNCTION
TEMPERATURE OF 100·C.
. JUNCTION TEMPERATURE IN ·C
FIGURE 19.3 ESTIMATED FAILURE RATE DF A STANDARD SILlCDN CONTROLLED RECTIFIER AS
A FUNCTION OF JUNCTION TEMPERATURE, REVERSE AND/OR FORWARD VOLTAGE,
AND DEFINITION OF FAILURE FOR A MAXIMUM RATED JU.NCTION TEMPERATURE
OF 1000 C
SCR MANUAL
,"
"~",
KEY
" ",
- - DEGRADATION DEFINITIDN OF FAILURE' "
- - - CATASTROPHIC DEFINITION OF FAILURE'
NOTES:
I. FOR ESTIMATING FAILURE RATES OF DEVICES
SUBJECTED TO STRESS SCREENING TESTS,
DIVIDE THE FAILURE RATE OIiTAINED FROM
THIS CURVE BY 10.
2. THIS CURVE IS TO BE USED FOR ESTIMATING
FAILURE RATES OF SILICON CONTROLLED
RECTIFIERS WITH A MAXIMUM RATED JUNCTION
TEMPERATURE OF 125·C.
0.00001 L-:2d:00","""~,,-....,.Jb--+.,..--;;6,...--....,J;,-----b.,...--I
JUNCTION TEMPERATURE IN ·C
FIGURE 19.4 ESTIMATED FAILURE RATE OF A STANDARD SILICON CONTROLLED RECTIFIER AS
A FUNCTION OF JUNCTION TEMPERATUREL REVERSE AND/OR FORWARD YOLTAGE,
AND DEFINITION OF FAILURE FOR A MAAIMUM RATED JUNCTION TEMPERATURE
OF 125°C
10.0r=--r--,--,---,---,---r---,---,
1.0
O.Ib--+--+--lA;"'~-~d'-.N""+---+-l
W
"'w
~LTIME
(b) Waveform of Anode Current and On-State Voltap
FIGURE 20.17
HIGH LEVEL ON-STATEVOLTAGE TEST
Figure 20.17(a) shows a block diagram of a circuit where the high
current pulse is generated by discharging a capacitor through an inductor. Operation of this circuit is as follows: A regulated dc voltage source
is used to charge capacitor C l to a specified voltage level. When the
initiate button is depressed, capacitor C l will discharge through SCRb
L l , D.U.T. and Rs. The discharge current waveform will be a halfsine wave, with the base width determined by the following formula:
580
TEST CIRCUITS FOR THYRISTORS
T = time is seconds
F= resonant frequency of C 1 & Ll in Hz
The resonant frequency is:
F
FinHz
Lin Henries
C in Farads
1
= ---__:___==::_
2·
7r.
YL·
C
The peak current:
I peak in Amps
Em volt on C 1 in Volts
Z impedance of discharge path at resonant
I peak = Z
frequency in ohms
Z should be « VTM/lp or the waveform will not be sinusoidal. The
peak voltage which appears at D.U.T. will be measured at peak reading
storage voltmeter.
At the end of the forward current pulse the current, due to resonance attempts to reverse (the circuit being oscillatory). SCR1 will prevent this, but the resultant reversal of the voltage across SCR1 does
insure its tum-off (tq < T).
Capacitor C 1 will now slowly recharge through resistor R1 • The
gate signal to fire the D.U.T. is derived from the anode voltage source.
An independent gate signal is sometimes used instead where synchronization of this signal is important. SCR2 is in the circuit only to protect
the operator from high voltages which may appear on the test terminals. This could occur if SCR1 is fired, but no D.U.T. is connected,
or if D.U.T. fails to fire.
Figure 20.18 shows an actual circuit to generate pulses between
0-125A with a base width of 1 ms.
Em
ADJUST
CURRENT
~II,---+--~
T,
T2
SNez TO SCOPE
-
FOR
-
CURRENT
ADJUST
TI
VARIABLE TRANSFORMER GE 9T92AI (2.5- 3.5A O__ 120__ 140Vl
T2
STANCOR PC-8301
CR I _2
R,
DIODE GE MPR 15
SK,IOW
C,
R2
.2 OHM 1'%.5W
STANCOR PC-8301 (830V CT 200mA)
90~F OIL.·CAPACITO~
GE 4
x
20~F
NO.960
I
x lo,..F NO- 959
SCRI_2 THYRISTOR GE C35N
eR!
DIODE AI4F
LI
1.lImH INDUCTOR MILLER 3 x .37mH NO.7827
CR4
R3
RG
ZENER DIODE == 40V IW
RESISTOR 2204. 2W
RESISTOR DEPENDS ON D.U.T.
FIGURE 20.1'
ON·YOLTAGE TESTER
581
SCR MANUAL
20.9 CRITICAL RATE OF RISE OF ON-STATE CURRENT TEST ,(Dl/DT)
(See JEDEC Rating Establishment and
Verification Tests - Part 5)
When thyristors are switched by a gate signal from the off-state
into a high on-state current, they tend to begin conducting in a limited
area physically near the gate contact. (This, of course, depends on the
type of gate construction employed.) This conducting area then spreads
with time until the entire area of the device is conducting. If the offstate voltage is high and the on-state current rises rapidly, it is possible
to dissipate a peak power of many kilowatts in a very smail area, during
the switching interval. This causes very high spot temperatures with
resulting high switching loss. Ultimate failure by spot melting through
the junction can occur in the extreme.
The di/dt test is a measure of the ability of the thyristor to withstand switching from high off-state voltages into fast rising load currents. The gate signal used is normally high in amplitude with a fast
rising leading edge, to make the initial conducting spot as large as
possible. Both the amplitude and rise time of the gate signal are
important.
Two different non-repetitive ratings may be assigned to thyristors,
one for gate triggering (which is described and most commonly used)
and one for triggering by exceeding the thyristor breakover voltage.
Current waveform and numerical value of di/dtare shown in
Figure 20.19(a) and (b).
...-'Ii'Vlr""""*:----tT---{OJ TO SCOPE
SI
= MERCURY WETTED REED RELAY OR SCR
RI
= NON INDUCTIVE
= CURRENT LIMITING
R2
RESISTOR
(c) Circuit Far Exponential dv/dt Test
FIGURE 20.22
dv/dt TEST (EXPONENTIAL)
Where this parameter appears on the device specification sheet,
it enables the circuit designer to design filters to prevent false triggering. Figure 20.22(c) illustrates a simple circuit to check thedv/dt
capabilities of thyristor devices.
The operation of the circuit is straightforward, but a few rules
have to be observed to obtain good results. The switch 51 can be a
mercury wetted relay or 5CR, but its closure time (including bounce)
must be less than 0.1 . Rl . C l . Resistor R2 is used to limit current in
the event of breakover. The values of Rb R2 and C l must be selected
to :minimize waveform distortion due to thyristor and circuit wiring
impedance. The rate of dv/dt is increased by lowering C l or R l . Rs
will discharge C l after 51 is opened.
20.11.2 Linear dvI dt Test
This test is performed with a linear waveform of specified ampli-
586
TEST CIRCUITS FOR THYRISTORS
tude with the device initially unenergized. The rate is increased until
the thyristor breaks over. The rate of rise at the breakover is the critical
value. The test voltage waveform and numerical value of the critical
rate of rise are shown in Figure 20.23
Jt!..
dt(LlN)
.O.8~
t2-tl
0.'-14_+--_ _ _ _ ___
TIME
(a) dv/dt Test Voltage Waveform (Linear)
(b) Numerical Value of dv/dt (Linear)
(e) Circuit for Linea, dv/dt Test
FIGURE 20.23
dv/dt TEST (LINEAR)
Figure 20.23(c) shows a basic circuit to generate a linear ramp.
Initially the low voltage, high current supply is circulating a current 11
in the loop R 2, Ll and Dl whose amplitude depends on the Supply 1
adjustment. Supply 3 will charge C a to a negative voltage through the
high impedance of R5 . After 11 has stabilized and C a is charged to a
negative voltage, SCR1 can be fired and constant current can flow
through Ra into Ca.
The voltage on C a will rise in a linear fashion from - Ea to +E 2.
The D.D.T. does not see the linear ramp until D2 becomes forward
biased, i.e., at about 2 volts above ground. The slope of the ramp can
be varied by adjusting the magnitude of the constant current and/or
adjusting the capacitance of Ca. The voltage amplitude to which this
ramp rises is determined by E 2, where El << E2 and E2 is programmable. The voltage waveform can be observed with a scope across the
D.D.T.
Slow turn-on of SCR 1 and slow reverse recovery of Dl can cause
non-linearities in the lower part of the voltage waveform, which is
eliminated by R5 and D 2. The reverse voltage on D.D.T. when SCR 1
is off should not be above 0.02 . (El + E 2).
SCR 1 should be on for a minimum of 50 p.Seconds, R6 is required
to prevent damage to the D.D.T., but should be as small as possible
to minimize waveform distortion. SCR2 assures turn-off of SCR1 in
case the D.D.T. turns on. Lead inductance and length in the loop
containing SCRlo Ra, Ca, D2 and R6 should be kept to a minimum.
587
SCR MANUAL
20.12 CRITICAL RATE OF RISE OF COMMUTATING OFF·STATE
VOLTAGE FOR BIDIRECTIONAL THYRISTORS (TRIACS) TEST
The bidirectional thyristor, in its usual mode of operation, is
required to switch to the opposite off-state polarity following current
conduction in the on-state. In common 50, 60 and 400. Hz AC phase
control applications utilizing sinusoidal voltage sources, this, switching
occurs each half cycle at the current zero point.
As long as voltage and current are in phase, the reapplied voltage
rises relatively slowly, but operation of a triac into an inductive load
requires additional design constraints that must be considered for
during circuit design. The most serious of these constraints is commutation dv/dt. It is largely influenced from the last 1h of the decreasing
current di/dt (for more details see Chapter 7).
Commutation dv/dt is that rate of rise of voltage impressed across
the triac by a circuit at the cessation of current How. In an inductive
AC circuit current lags voltage by a phase angle () and as a result current goes through zero sometime after the supply voltage has reached
a finite value in the opposite direction. Since the triac tries to turn off
at zero current, the instantaneous line voltage at that point appears
suddenly across the device at a rate limited only by circuit stray capacitance and the capacitance of the triac'itself (C T ). For the triac to turn
off reliably in an inductive circuit, the rate of voltage rise across the
device must be kept within specified limits, and the test circuit of
Figure 20.24 is intended to check this dv/dt withstand ability. In this
circuit, the rate of rise of voltage across the triac is made adjustable
(from 10 volts per microsecond down to less than 0.3 volts per p.Second)
by deliberately adding capacitance C 1 in shunt with the test device.
Resistor R2 prevents high peak current from Howing through the triac
when it turns on and discharges' C 1 •
--_=-::-::-::-::-::{o) rc't:~~:~)
.----.,9-~_--_--_-----=-
R,
'W
100.0
C,
IOOpF-Ip.F
VARIABLE
(a) Commutating IIv/dt Test For Triacs,
COMMUTATION
OF TRIAC
(II) Suppl, Voltage & Loa!! Current
(e) Rate of Ri'e of Voltage After
Triac Commutates
(d) Expansian of
Commutatlng dvldt
FIGURE 20.24
588
COMMUTATING dvldt TEST FOR TRIACS
TEST CIRCUITS FOR THYRISTORS
Design Equation
The circuit shown in Figure 20.24 can be used to test a triac with
a current rating of I RMS = 10 amps - see Table 20.1.
di/dt = 6.28 . f . ITM • 10- 3
(Alms)
I RMS ' y2 6.28·60 . 10·1.41
5.3 Alms
=
ZL
==
=
fI)'
=
= 115· y2 = 11.5 n
EM
10· y2
ITl\l
ZL = yXL2 + R2
Choke selected as 30 mH
XL
L
2 . 3.14 . 60 . 30 . 10- 3
=
fI)
•
=
XL => 10
Recommended: Ii;
= 11.31 n
R3 = 1 n
An on-state current duration of 90% of a half cycle is recommended,
which means that heat sinking is required.
Testing Procedure
Testing procedure is as follows: Set C 1 initially to 1 pi and Rl to
maximum resistance. With power applied adjust Rl to obtain 90%
on-time and 10% off-time.
U scope is connected as shown in Figure 20.24(a), you should see
the current and voltage as shown in Figures 20.24(b) and (c). (Voltage
is 180 0 inverted.)
U testing has to be done at elevated temperature provision for
external heating and monitoring of case temperature (junction temperature) has to be provided. C 1 is then progressively reduced until the
desired rate of voltage rise across the triac is reached or failure to
commutate results as monitored by the test scope.
Numerical rate of voltage rise is defined by the waveforms of
Figure 20.24(d).
Type
I
Rating
di/dt
50Hz
di/dt
di/dt
60Hz 400Hz
ZLin
Ohms
SC35/SC36
SC40/SC41}
SC240/241
3
1.33
1.6
6
2.66
3.2
21.5
19.2
SC45/SC46l
SC245/246
lO
4.5
5.4
36
11.5
SC50/SC51
SC250/251
SC60/SC61
SC141
SC146
15
25
6
10
6.66
11.2
2.66
4.5
8
13.5
3.2
5.4
54
7.7
89
4.6
TABLE 2G.1
38.4
dl/dt VAlUES FOR MOST COMMONtY USED TRIACS FOR 50, 60 AND 400 Hz
20.13 TURN-OFF TIME TEST
As discussed in Chapter 5, the turn-off time depends on a number
of circuit parameters. Thus, a turn-off time specification inherently
589
SCR MANUAL
must include the precise value of these circuit parameters to be meaningful. Accordingly, specifications for General Electric SCR's with
guaranteed turn-off limits list the applicable circuit parameters and
show the test circuit which will apply these parameters to the SCR.
For this information, the reader is referred to the specification bulletin
for the SCR under consideration.
For general turn-off time test work, the type of circuit shown in
Figure 20.25 can be used for low, medium, and high current SCR's by
proper manipualtion of the circuit constants. Forward load current is
adjustbale by R5 from approximately 1h ampere to 70 amperes and the
length of time SCR 1 is reverse biased during the turn-off cycle can be
adjusted by manipulating R5 and C 1 . The peak reverse current during
the recovery period .can be adjusted by R7 and this current can be
viewed on a scope by monitoring the voltage across a non-inductive
shunt Rs. If one is interested in only one particular load current range,
there is of course no necessity in providing the complete range of
resistors and capacitors specified in Figure 20.25.
FACTOIn'TESTa
GUARlHTEEDLII/IIT
~/Ir---'~''''
r-"'OLTS
I
~/I
I
RI ,R 3 -IOO OHMS,IWATl
RB-NON-I~OUCTIVE S~UNT
SCRITURN-OFFTIME
OCRlTtlRN-OfFT~'1
>"
(IF USED)
R~,R"_I MES,IWATT
CI-,~6'~g::f~d PAPER CAPACITORS,
A5-~A5R~~B~ ~~~'.~~~~frn,
C2,C3.-0
1l6-IOOOIl,2WATT
SCRI-TE5TCON~OLLEORECTIFIEl'I
R7 -O.I£I,500W TOlil,lOWATT
5<;R2- Gr C22F
5f"d-ZOO~OLT
I
/
PAPER CAPACITORS
FIGURE 20.25
TURN-DFF TIME TEST
This test circuit subjects the SCR to current and voltage waveforms similar to those found in a parallel inverter circuit. Closing Sl
and S3 fires SCRb the unit under test, so that load current flows through
R5 and the ammeter. In less than a second C 1 charges through Ra to
the voltage being developed across Ro by load current flow. If S2 is
now closed, SCR 2 turns on. This applies C 1 across SCR1 so that the
current through SCR1 is reversed. C 1 furnishes a short pulse of reverse
recovery current through SCR 1 until this SCR recovers its reverse
blocking ability. After this initial pulse of current, C 1 continues its discharge through SCR2 , the battery, and R5 at the rate dependent on the
time constant of RoC l . After a time interval, tl> in Figure 18.7, somewhat less than the RoC l time constant, the anode to cathode voltage
of SCR1 passes through zero and starts building up in the forward
direction. If the turn-off time, t"ff, of the SCR is less than tl> it will
remain turned off and the ammeter reading will return to zero. If not,
the SCR will turn back on and current will continue to flow until S3
is opened.
590
TEST CIRCUITS FOR THYRISTORS
The tum-off interval t1 can be measured by observing the anode
to cathode voltage across SCR1 on a high speed oscilloscope. A waveshape similar to that shown in the figure will be observed.
Satisfactory operation of this circuit requires careful attention to
detail. The DC source must have good regulation if C 1 is to develop
ample commutation voltage for turning off SCR1. In order to minimize
circuit inductance, power leads should be heavy copper braid when
testing medium and high current SCR's and lead lengths should be
held to an absolute minimum.
Tum-off time testing in the General Electric factory is performed
with a fixed rate of rise of reapplied forward voltage as indicated by
the dashed line in Figure 20.25. This is a more severe test on the SCR
than the exponential curve and it requires considerably more elaborate
test equipment than in Figure 20.25. For those who wish to test under
these conditions, information on the factory test circuit will be provided upon request. (Ref. 5.)
Tum-off time is sometimes specified with an inverse diode connected in parallel with the D.U.T., which is a more severe test. Special
attention should be given to reverse bias conditions when comparisons
of "tum-off time" are made.
20.14 THERMAL RESISTANCE TEST,
The thermal resistance of a thyristor is a measure of the ability
of the thyristor package to remove heat from the silicon pellet. Therefore it is a limitation on the power handling capability of the device.
Thermal resistance is measured in degrees Centigrade per watt, that is,
in degrees Centigrade of temperature rise for each watt of power
dissipated.
Since it is impossible to measure the junction temperature rise
directly, the temperature dependence of the on-state voltage with a
low-level current Howing is used to measure junction temperature rise,
while a constant power is being dissipated under constant cooling
conditions.
A thermal resistance test set must supply the on-state current (heating source), the reference current supply and an on-state voltage drop
measuring circuit. A heat sink must be provided for the test device.
A simplified version of a practical thermal resistance tester is shown .in
Figure 20.28 which can be used for SCR or rectifier measurements.
The measurement of thermal resistance, junction to case, consists
of making measurements to satisfy the following equation:
TCl- T 02
R
_ TJ-To
em - P(AVG)
VT(HTG) . IT(HTG) . Duty Factor
T Cl
The measured case temperature with only metering current Howing
This is measured case temperature when the
T 02
thyristor is mounted to a heat dissipator and
operated with power applied
IT(HTG) = Heating current
VT(HTG)
The measured value of on-state voltage when
IT(BTG) is applied
=
=
=
591
SCR MANUAL
When thermal resistance, junction to ambient T C2 should be replaced
by TA (ambient temperature).
Test Procedure - Step 1
First the D.V.T. is operated with power intermittently applied,
but at very high duty cycle. During the intervals between power pulses,
the heating current is removed and with metering current Bowing, the
metering voltage is measured.
The D.V.T. current and voltage waveform are shown in Figure
20.26(a) and (b) for a 60 Hz repetition rate.
----ITlHTG
I
I
I
I
I
o
_ __
~,W,~t:===4J~--~-~:I~T~(M~E~T~I
__~====::~--II
I
,,
II,
II
1'2 '3
'4'
tl
tl5
"I
I
:
'4 -'I" 0.333 MILLISEC MAX
:IT(MET)=METERING CURRENT
"1-'
IT (HTG I ' HEATING CURRENT
I' 16.7 MILLISEC
DUTY CYCLE =0.98 MIN.
VT(HTGI
=HEATING
VOLTAGE
(al Current Waveform
~METERING
1
i
I
I
INTERVAL
1
HEATING VOLTAGE LEVEL
V T It I IS THE EXTRAPOLATED
I METERING VOLTAGE AT 'I
I I
I I
I I
I I
I I
V
I
THII-I
- - - - - A C T U A L METERING VOLTAGE WAVEFORM
I
'I
'2
'3
(bl VDltage Waveform
FISURE 20.26
CURRENT .. VOLTAGE WAVEFORMS DURINS THERMAL RESISTANCE TEST
The metering current which Bows continuously must be held
constant. This is particularly important during the metering interval
between power pulses, because the test device impedance will vary
considerably during that time.
It would be desirable to arrive at the thyristor virtual junction
temperature at the exact instant when the heating current removal is
initiated since the virtual junction temperature will be maximum at
that time. However this is not possible. First it takes a finite time for
the thyristor current to decay from the heating current value to the
metering current value (t2 - tl in Figure 20.26(b). Secondly, transients
will exist in the metering voltage waveform for some time after the
metering current value is reached due primarily to charge storage
effects in the thyristor. The time ta on the waveforms rep.l~esents the
592
TEST CIRCUITS FOR THYRISTORS
shortest time after removal of heating current that metering may be
measured. For a particular device type the time ta is best found by
performing the test at various power levels and noting the shortest time
where the measured value of thermal resistance is essentially independent of the power dissipated. Power levels of 25% above and below
the power corresponding to the specified heating current are recommended for this determination. Time ta should be expected to be in the
range of 100 to 200 microseconds.
Since some active element cooling occurs between the time when
the heating current is removed and time t a, the thermal resistance value
determined from a metering voltage measurement at ta will be in error,
it is therefore necessary to extrapolate the metering voltage waveform
back to tl from ta based on the shape of the waveform from ta to t4
where the waveform is a true representation -of the junction temperature cooling curve. An exponential curve is a reasonably good approximation of the true cooling curve. In the time range of interest, the
exponential curve is nearly linear because the exponential time constant of the device cooling curve is relatively long. Therefore, linear
extrapolation of the actual cooling curve from time ta back to time tl
results in little error and is recommended. Figure 20.26(b) illustrates
the extrapolation.
Step 2 - Determination of Junction Temperature
The power application test (Step 1) produced a value of on-state
voltage at the metering current level which corresponded to the maximum virtual junction temperature attained. Step 2 consists of operating
the test device with no significant power dissipation so that for all practical purposes the thyristor virtual junction temperature and case
temperature will be equal. The thyristor is operated at the same value
of metering current as in Step 1. The on-state voltage is monitored and
the thyristor is externally heated on a temperature controlled block or
in an oven until the measured value of on-state voltage equals the
extrapolated value VT(tl) obtained previously. When the on-state voltage has stabilized, the thyristor case temperature is recorded. This
value is T Cl •
When the metering current is initiated for Step 2 of the test, it
should momentarily be increased to the value of IT(HTG) used in Step 1.
This is to assure that the device is fully turned-on. The duration of the
IT(HTG) pulse should be at least one.millisecond but not longer than
five seconds to avoid unnecessary heating of the test device.
If a continuous gate current value is used as a test condition for
Step 1, it must also be used in Step 2.
593
SCR MANUAL
r - ---
II~
60 Hz
I
--1
~~~NT
OC- POWER
..
GOV
..,1-------~-~____1t__------____,
I
I
L ____ .J+
r----l
115v-l1
I
TRIGGERING
60 Hz
CIRCUIT
R,
C,
seR2
i--i:=====:::::~
CR.
--1L ____ _Ji
CR
3
115~
r----..,L . -
6V I
~j::~~~ I
MIN I ADJUST VOLT.
DC-POWER IINPUT
MAX :~PLES L - -
CR 4
METERING
I
CURRENT
I
6D ~ ADJUSTABLE 1_
+
I
i2!.--
-L ____ J
TO
DIFFERENTIAL
COMPARATOR
SCOPE
L ____ .J~--------------~--------~---J
FIGURE 20.27 THERMAL RESISTANCE TEST CIRCUIT
Test Circuit
A basic circuit which may be used for testing the thyristor in
Step 1 with high level (heating) current present is shown in Figure
20.27. The active element of the D.U.T. is heated by direct current
having an rms ripple content of 5 percent or less which is passed continuously through the D.U.T. except for a 0.333 millisecond maximum
interval every 16.7 milliseconds. During this 0.333 millisecond period,
the junction temperature is indicated by reducing the on-state current
to the metering current value and measuring the on-state voltage. This
circuit will produce the current and on-state voltage waveshapes. shown
in Figures 20.26(a) and (b).
Control of the heating current through the D.U.T. is accomplished
by SCRI and SCR2 (see Figure 20.27) which functions as a dc flipHop switching at a 60 Hertz repetition rate to facilitate oscillographic
observations. Current is carried by SCRI only during the on-state voltage metering interval so this SCR may be considerably smaller than
SCR2. Cl> which is charged by the low current dc power supply has
the function of turning off SCR2 when SCRI is triggered.
Unavoidable inductance in the heating current power supply and
associated circuit wiring make it impossible to turn off the heating
current abruptly without creating transient voltages which would interfere with the measurement of on-state voltage. To overcome this, a
diverter circuit consisting of rectifier diodes RD 1 through RD5 is
included so that heating current is not interrupted by SCR2, but is simply switched to a different path. The inductor L may be included to
make certain that the heating current does not vary while it is being
switched from one path to the other. This inductor also serves to reduce
to a negligible amount undesired How of current from C 1 through the
D;U.T. and the heating current power supply. The inductance in the
diverter circuit should be kept low so that 10 !.tS after SCR 1 begins to
594
TEST CIRCUITS FOR THYRISTORS
contact, all heating current will have been diverted away from the
D.U.T. In Figure 20.27 the portion of the circuit in which inductance
must be carefully controlled is indicated by heavy lines.
In order to observe the on-state voltage of the D.U.T. during the
metering current interval, the use of a differential comparator preamplifier is recommended.
20.14.1 Thermal Resistance of Press Pak
Rectifier Diodes &Thyristors
The Press-Pak configuration makes possible a very simple technique for measuring thermal resistance of either rectifier diodes or
thyristors without the complications and inaccuracies associated with
the junction ..temperature measurement. Since there are approximately
two equal heat flow paths from junction to ambient, heat can be passed
through the device from an external source to a heat sink. The heat
flow can be measured and the heat flow divided into the temperature
drop across the device giving the thermal resistance of its two heat
flow paths in series. This method is explained in detail in Reference 1
at the end of the chapter.
20.15 TESTING THYRISTORS ON CURVE TRACERS
Curve tracers, like the Tektronix 575 and 576, are well known
instruments for measuring diodes and transistors. They are also very
useful to measure thyristor characteristics .
.for f\
,......----,
OR NV\OR
\Tv
\liN
COLLECTOR
SUPPLY
VERTICAL
DEFLECTION
PLATES
DATA FOR A 576 CURVE TRACER:
COLLECTOR SUPPLY
O___ 1500V
OR O.IA ___ IOA
PULSED 20
HORIZONTAL DEFLECTION
0.05 V/CM ___ 200VlCM
STEP GENERATOR
5mA ___ 2A
OR 5mV ___ 40V
ImA/CM ___ 2A/CM
VERTICAL DEFLECTION
PULSED MODE FOR COLLECTOR SUPPLY
AND STEP GENERATOR ARE POSSIBLE
FIGURE 20.28
BLOCK DIAGRAM OF CURVE TRACER CONNECTED TO THYRISTOR
The block diagram and typical data for a Tektronix 576 in Figure
20.28 shows the suitability for testing some of the characteristics of
thyristors.
595
SCR MANUAL
. 20.15.1 Off·State & Reverse Voltage
CURRENT IDlY
I
~T2
IC)
ANODE
IC)
....._+----~~__ YOLTAGE/DIY
_+-~
-----
IRM
~~'"'"
IB)
(B)
GATE
GATE
IE)
(a) Scope Display of Off-5tate & Reverse
Voltap Test
FIGURE 20.29
.',
IE)
(II) Connection of Thyristors to
Curve Tracer
DISPLAY ON CURVE·TRACER ANO THYRISTOR CONNECTIONS
The device is connected as shown above, the gate terminal is
returned to the cathode through a resistor if specified in the specification sheet. Then the voltage between anode and cathode is increased
:and the leakage current can be seen as the vertical deflection.
Figure 20.29 shows a scope display where an AGvoltage was used
to measure simultaneously the off-state and reverse directions.
20.15.2 Gate Voltage, Gate Current Measurement
After the thyristor is connected as shown in Figure 20.29(b), the
collector supply is adjusted to 12 volts (or whatever anode voltage
required, but keep in mind that gate voltage and current to trigger are
functions of anode voltage and junction temperature). The proper
anode resistor can be selected (dissipation limiting resistor) and the
proper gate-cathode resistor could be connected externally if necessary.
The step selector should be set to minimum current per step.
Now the D.U.T. is connected to the test circuitry of the curve tracer.
The horizontal amplifier will display the anode voltage (12 volts). This
amplifier can then be switched to a position which will connect it to
the step-generator. Depending on the position of the step generator
(voltage or current) either parameter could be displayed as a horizontal
deflection.
IbiZ
ow
00:
ZO:
-c,o
TRIGGER POINT---....
4
Imo ID1V
(a) Curve Tracer Display of Gate
Current to Tngger
596
TRIGGER POINT
00:
zo:
"a
3
FIGURE 20.30
IbiZ
Obi
.1
.2 .3
.4 .5 .6 .7
O.IY IDlY
(II) Curve Tracer Display of Gate Voltage to
Trigger Series Resistor = 0
GATE VOLTAGE & GATE CURRENT DISPLAY AT CURVE TRACER
TEST CIRCUITS FOR THYRISTORS
Figure 20.30(a) and (b) show gate current and gate voltage measurement. The sensitivity at every step depends on the setting of step
selector switch. By using steps/family and step zero adjustment reasonably accurate measurements can be made.
Another possibility is to display gate current and gate voltage
simultaneously.
~
0.1'
TRIGGER
POINT
;! 0.6
g
0.&
~
0.4
~0.3
~ 0.2
dO"
t
2
:5
..
!5
6
I inA/DIY GATE CURR[HT Im,.l
FIGURE 20.31
GATE VOLTAGE ANO GATE CURRENT TO TRIGGER FOR THYRISTOR,
DISPLAYED ON CURVE TRACER
The horizontal amplifier has to be switched to display base current.
Sensitivity can be selected by the step selector switch (rna/step) and
the vertical amplifier is switched to "base volts" position, which connects this amplifier to the D.U.T.'s gate cathode terminals. The series
gate resistor no longer influences the gate voltage measurement.
20.15.3 Forward Current & On-State Voltage Measurement
The Tektronix 575 and 576 can be used to measure the on-state
voltage of power thyristors and triacs up to 10 amps and in a pulsed
mode up to 20 amperes. (The 576 pulsed high current fixture increases
the step generator and collector supply by a factor of 10.) This is sufficient for low and .medium current thyristors. After the D.U.T. is connected as shown in Figure 20.29(b), a low voltage and an appropriate
series resistor on the collector supply is selected. The horizontal amplifier will display the anode current and the vertical amplifier will display
the on-state voltage of the device. The device is triggered into conduction by increasing the gate current on the step selector switch. Anode
current can now be increased by increasing the collector supply voltage or decreasing the dissipation limiting resistor. The horizontal deflection will allow a convenient reading of the on-state voltage of the device
at the appropriate current.
+YON
-YoN
---7'1~--
...
(8) Thyristor
FIGURE 20.32
(b) Triac
DISPLAY ON CURVE·TRACER OF ON-5TATE VOLTAGE MEASUREMENT
597
SCR MANUAL
More measurements like holding and latching current measurements can be done. For'more information see references 2, 3, and 4.
20.16 ELEVATED UMPERATURE TESTING
In Chapter 12, Zero Voltage Switching, there is a wealth of information on temperature controllers which would be very suitable for
heating the test .devices wherever elevated temperature testing is
necessary.
20.17 COMMERCIAL THYRISTOR TEST EQUIPMENl'
Several manufacturers offer ready-made thyristor test equipment,
or design and build it to customer specifications. Some are listed below
and should be contacted directly for details.
Cyberex Inc.
4399 Industrial Parkway
Willoughby, Ohio 44094
Utah Research & Development Co., Inc.
1820 South Industrial Road
Salt Lake City, Utah 84104
Tektronix, Inc.
P.O. Box 500
Beaverton, Oregon 97005
Mastech Inc.
478 East Brighton Ave.
Syracuse, New York 13210
REFERENCES
1. "Pressure Contact Semiconductor Devices," W. Warburton, W. F.
Lootens, T. Staviski, IEEE IGA Conference Recor~ 1966.
2. Semiconductors Device Measurements, First Edition, Tektronix,
1968.
3. Tektronix Instruction Manual, Type 575 Curve Tracer.
4. Tektronix Instruction Manual, Type 576 Curve Tracer.
5. "Turn-Off Time Characterization and Measurement of Silicon Controlled Rectifiers," R. F. Dyer and G. K. Houghton, AIEE CP
61-301. (Available as General Electric Application Note 200.15*)
6. JEDEC Recommended Standard for Thyristors, available from Electronic Industries Association, 2001 Eye Street, N.W., Washington,
D.C. 20006.
*See Chapter 23 for ordering information.
598
SELECTING THE PROPER THYRISTOR AND CHECKING THE COMPLETED CIRCUIT DESIGN
21
SELECTING THE PROPER THYRISTOR AND
CHECKING THE COMPLETED CIRCUIT DESIGN
21.1 SELECTING THE PROPER THYRISTOR
A glance at the device specification section in Chapter 22 shows
that the equipment designer has available to him a wide range of
thyristor components from which to choose. Basic SCR types are·
offered with current ratings extending from 0.8 amp to 1400 amps
RMS, and with voltage ratings spanning the range 15 volts through
2600 volts peak. In many instances, within this range, economy/light
industrial SCR's exist side by side with similarly rated industrial military types. Many specialized SCR types also are listed, including high
speed inverter SCR's with guaranteed dynamic characteristics, SCR's
for use over very wide temperature ranges, SCS's, PUT's light-activated
SCR's, very high voltage SCR's, and SCR's tested to very rigid quality
levels for high reliability applications. Bidirectional thyristors (triacs),
intended primarily for use on 120 volt and 240 volt AC power lines,
are presently available in 3, 6, 10, 15 and 25 amp sizes. Diacs and
UJT's, while not strictly speaking thyristors, are included because of
their wide usage as thyristor trigger components. Packaged assemblies
("stacks") of individual thyristors and/or rectifier diodes-both with
and without suitable control circuitry-complete the range. For the
equipment designer understandably confused by this profusion of
types, the following selection criteria are offered.
21.1.1 Semiconductor Design Trade-Offs
Within the present state of the power semiconductor art, it is true
to say that there is no such thing as a "universal thyristor." An SCR
optimized for use in a high speed inverter or chopper circuit for
instance, may be a bad choice for use in a 50 or 60 Hz phase control
application. By the same token, a thyristor designed for use in very
high voltage applications is by nature unsuited for use in high frequency circuits. These various incompatibilities stem from the fact that
most device design approaches leading to good high power handling
capabilities (voltage or current) are diametrically opposite to those leading to good high frequency performance. As a result state of the art high
frequency devices tend to have limited power handling capabilities,
while the highest power devices are relatively slow. Between these two
extremes· there are naturally many general-purpose devices that combine medium speed performance with medium power handling capabilities. Figure 21.1 summarizes some of the design factors that affect
practical thyristor electrical performance at this writing.
599
SCR MANUAL
EFFECT ON
Desip Varlalile
(Increase)
CUrrent
Ratlq
PelletArea
(emitter)
+-
Voltage
Ratiq
Base Width
~
Resistivity
oj,
Lifetime
...
...
...
...
'"
...
Thermal Resistance
Surface Contouring
oj,
Emitter Shorts
Optimized Gate
Structure
~
oj,
...
Turn Off
Time
IIv/dt Withstand
AIIlllty
Allillty to Switch
HIIb Currents
RatldlJ
(d/dt)
~
~
+
+
~
...
+
oj,
't
...
~
(See Chapter 1 for Discussion)
Key: Beneficial EffectsIncrease ...
Decrease 't
FIGURE 21.1
Undesirable EffectsIncrease
Decrease oj,
+
THYRISTOR DESIGN TRADE-OFFS
Chapter 1 contains more information on these different design
trade-offs. One critical item is the gate structure. The simple point gate
is satisfactory for low di/dt applications but more intricate gate designs
are required as di/dt stress increases. These latter designs sacrifice
emitter area so that the current rating decreases for a given silicon
pellet size.
There are also several design compromises discussed in Chapter 1 that can be made in the mechanical construction of a thyristor.
For example, when a thyristor is designed specifically for use in the
light industrial and consumer markets-environments characterized by
limited temperature excursions and absence of wide range cyclical
loading-simple low cost fabrication techniques are usually employed
in its construction. Such techniques, while completely adequate for
their intended purpose, would be completely unacceptable if applied
to the design of a 500 amp SCR destined for use in a steel mill drive.
Here, a premium thermal-fatigue resistant and high voltage· structure
would be a "must."
21.1.2 Selection Check List
To select and apply any thyristor successfully, none of its published ratings should be exceeded. Equally evident, it would be uneconomical to apply the device too conservatively. To make a proper
device selection then, the equipment designer should first of all prepare
a check-list outlining all the limiting conditions of his particular application. Section 21.4 contains all of the component specifications that
have to be considered. Since thyristor ratings are usually specified as
maximum or minimum values (worst case), the designer subsequently
can determine which device best !its the needs of the application. The
following is a check list of the steps that should be taken or considered
in selecting the proper thyristor for a given application.
600
SELECTING THE PROPER THYRISTOR AND CHECKING THE COMPLETED CIRCUIT DESIGN
Step 1. Determine Circuit Requirements on Thyristor
Voltage across and current through the thyristor must be determined in terms of circuit input voltage and output power requirements.
Figure 9.4 shows these relationships for some common SCR circuits.
Note. Check voltage transients (Chapter 16).
Check current carrying capability required of the thyristor if the
current waveform is irregular (Chapter 3) or has a high starting
component.
Determine temperature range over which the circuit must operate.
Is a high-reliability or "MIL-Spec" device desirable or mandatory?
Step 2. Select Proper Thyristor
Refer to Section 22.1 and then to the more detailed specifications
in Chapter 22. For final check, consider individual device specification sheets with more detailed information.
Step 3. Determine Proper Heatsink
(a) Check maximum allowable ambient temperature if a lead
mounted device was selected.
(b) Select proper size heatsink from fin curves given on the specification sheet for stud mounted types. OR determine the power dissipation of the device in order to design an air or liquid cooled heat
exchanger following Chapter 18. OR select suitable pre-assembled
thyristor stack assembly from the many types available as indicated
in Chapter 22.
Step 4. Design Triggering Circuit
See Chapter 4 for thyristor triggering requirements and design
criteria. Commercially packaged triggering circuits are also available
using magnetic, light sensitive or semiconductor components.
Step 5. Design Suitable Overload Protection, if Requir.ed
Protect the thyristors and associated semiconductors against short
circuit and other fault conditions'! In some applications, economic
factors and industry practice may preclude or not require protective
circuitry coordination. Do not overlook "normal. overloads" such as
cold inrush to incandescent light bulbs 2 or starting current of induction
motors, etc.
Beyond these elementary steps, there are often other considerations meriting special attention:
1. Series or parallel operation of individual thyristors-Chapter 6.
2. Radio interference suppression-Chapter 17.
3. Frequency response-Chapter 3 and Chapter 5.
21.2 CHECKING CIRCUIT DESIGN
The purpose of this section is to aid the designer in diagnosing
and curing poor performance in his completed circuit. It also provides
a step-by-step procedure for checking the design to ensure long life and
reliable operation of the thyristors.
601
SCR MANUAL
21.2.1 Thyristor Ratin.gs and Characteristics
Thyristors must be operated within their ratings as given in the
specification sheet. Do not design around samples; the sample may
well be much better than the type number would indicate. If production quantities are later involved, some thyristors may be received
which are, for example, of lower voltage capability than the sample or
they may have longer tum-off times, lower dv/dt's, etc. Use specification sheet limit values not data gleaned from samples.
Voltage and current measurements must be made on all thyristors
in the prototype. For this purpose an oscilloscope is essential. It should
have a rise time of less than 100 nanoseconds in order that the waveforms may be reliably scanned for steep wave fronts.
Measurements should be made under extreme as well as normal
load conditions. Include open-circuit operation, momentary overloads,
and the first starting cycle.
21.2.2 Voltage Measurement (See also Chapter 20)
Make sure that the probe is adjusted to give a Hat response. Make
sure too that no ground-current loops are present; the rule is that only
one ground lead should run from the circuit to the oscilloscope.
21.2.3 Current Measurement (See also Chapter 20)
Current measurements are more difficult to make accurately than
voltage measurements. No universal instrument is available but satisfactory results are obtained using a combination of the following types.
Current Probe. This is a clamp-on type of current transformer with
the secondary connected to an oscilloscope. An example is the Tektronix Type P6016 current probe, Figure 21.2. When used with an
amplifier this probe can handle 15 amperes peak to peak and has a
frequency response extending from 50 Hz to 20 mHz. It is especially
useful for measuring gate-current pulses because the readings are free
from external pick up.
FIGURE 21.2
602
.CURRENT PROBE AND AMPLIFIER
SELECTING THE PROPER THYRISTOR AND CHECKING THE COMPLETED CIRCUIT DESIGN
This type of instrument cannot measure DC and is liable to saturate if the DC component exceeds 0.5 ampere. The current range of
the current probe of Figure 21.2 may be extended by winding a current transformer as shown in Figure 21.3.
CURRENT TO
......-BE MEASURED
FIGURE 21.3
METHOD FOR EXTENDING CURRENT RANGE OF OSCILLOSCOPE CURRENT PROBE
The core may be of ferrite, powdered iron or powdered molybdenum
(typically Arnold Mfg. Co. Cat. #106073-2). The number of turns
current ratio, thus Figure 21.3 shows a 10:1 arrangement.
Tektronix now has another clamp-on current probe that can
measure both AC and DC. The Tektronix P6042 current probe is
designed for use with oscilloscope systems having either 50 ohm or
high-impedance inputs. The maximum currents it can measure depends
upon frequency and varies from 20 A P-P at 0.1 Hz to 2 A P-P at
50 MHz.
Current Shunt. The current shunt must be a non-inductive resistor
which is inserted in the circuit. The voltage across this resistor is then
observed on an oscilloscope. An inexpensive form of a current shunt is
described in Chapter 20 along with construction details.
A much more elegant design is shown in Figure 21.4. This shunt,
made by T & M Research Products, 129 Rhode Island, N .E., Albuquerque, New Mexico 87108, has a frequency response from DC to
150 MHz and can carry 60 amperes rms continuously. The only limitations of this form of current measurement lie in the practical difficulty
of inserting the shunt in the circuit and in avoiding false readings due
to stray pick up from ground loops.
=
FIGURE 21.4
A COMMERC IAL NON·INDUCTIVE CURRENT SHUNT
603
SCR MANUAL
21.2.4 The Power Circuit
The following anode voltage and current relations should be
measured on all thyristors in the circuit:
Peak forward blocking voltage
Peak reverse voltage
dv/dt
Turn-off time (tq) (if required, as in an inverter or chopper)
Rate of change of turn-on current (initial dildt)
Forward current before turn-off (if required)
Peak reverse current (if required)
Initial start-up current, e.g., inrush or latching, and holding
currents (if required)
Fault currents (if required)
These items are generally detailed in the specification sheet. If the
thyristor is running outside of specifications, either choose another
device with an improved rating or modify the circuit so as to run the
device within ratings.
21.2.5 Modifications to Soften dv/dt
Add a series RC network across the thyristor. Note that this may,
with low values of R, increase the di/dt. The effectiveness of the network may be increased by shunting a fast recovery diode across the
resistor. This increases softening of the dv/dt without worsening
the initial dildt (see Chapter 6).
21.2.6 Modifications to Soften Initial dildt
The initial di/dt may be limited by means of a reactor or saturating reactor connected in series with the thyristor. The design of the
saturating reactor is discussed in Chapter 5.
21.2.7 Gate Circuit
The following gate voltage and current relations should be measured in the prototype:
Gate voltage before triggering
Peak gate triggering voltage
Pulse width of triggering gate voltage
Gate triggering current
Gate current rise time
From the above data check that the following are within the
specified limits:
Peak and average gate power
Peak reverse voltage on gate
Peak gate triggering voltage
Note that for short trigger pulses the peak gate voltage that will
trigger all thyristors has to be increased as the pulse. width decreases
(Chapter 4).
604
SELECTING THE PROPER THYRISTOR AND CHECKING THE COMPLETED CIRCUIT DESIGN
Remember that a slowly rising gate pulse that will only just trigger
a thyristor is liable to increase local junction heating if fast rising anode
currents exist. Always trigger an SCR used in inverters with as steep
a rise time as possible (preferably shorter than 500 ns) and with as
high an amplitude as is permitted. Although this "hard" drive is not
always a condition of specification for some of the newer amplifying
gate SCR's, it never hurts to trigger hard (within rating) as the turn-on
is considerably better yet with hard drive.
Negative gate bias voltage may be applied to some SCR's that do
not have emitter shorting in the off-state to improve dv/dt and turn-off
time. This also eliminates random triggering due to noise. As always,
the data sheet must be checked to make sure that negative gate voltage
does not increase off-state blocking losses excessively (Section 4.3.5) or
worse yet, trigger on the SCR.
Where the anode current of an SCR is liable to oscillate due to
resonance in the load, it will be necessary to trigger the SCR with a
broad pulse. A gate pulse which did not extend to time t2 in Figure
21.5 would result in only the shaded part of the anode current Bowing.
By continuing the gate pulse to time t 2, the SCR will be retriggered
when the circuit again causes anode current to Bow. Extended gate
pulse duration is also necessary when triggering is initiated before
current zero in phase control applications with lagging power factor
load as discussed in Section 9.6.
II
.p
'"w
200
100 ~-H'-IIAI__+-------+-------I
b. Average Forward Power
Dissipation For Sinusoidal Current
Waveform
100
200
300
AVERAGE FORWARD CURRENT-AMPERES
FIGURE 21.12
ALLOWABLE CURRENT AND POWER DISSIPATION CURVES FOR THE C350 SCR
At 120° conduction 84 A average, the maximum case temperatures
allowed for these two devices are:
To (C3S0) = 93°C (double side cooling)
To (C180) = 100°C
The respective power dissipations are:
P (C3S0) = 160 watts
P (C1BO) = US watts
Determine the maximum permissible heatsink temperature under the
overload condition assuming that the C3S0 has a thermal contact resistance of .03°C/watt while that of the C180 is .08°C/watt when coated
612
SElECTING THE PROPER THYRISTOR AND CHECKING THE COMPLETED CIRCUIT DESIGN
with a thin layer of silicone grease.
T H.S. (C350) = 93 - (.03 X 160) =- 88°C
T H.S. (C180) = 100 - (.08 X 115) =- 91 °C
The final step is the heatsink design. The required thermal resistance of the heatsink to allow the SCR to operate in a 40°C ambient is:
Ts-TA
ResA (C350) =
PD(M)
88 - 40 =- 30 0 C/W
160
.
ReSA (C180)
=
91 ;;540 =- .44°C/W
Both SCR's would be reasonable choices. The final selection will
be economic since now the lower cost of the C350 is going to be traded
off against a larger and more expensive heatsink. Furthermore, protection for the C180 will be easier because of its larger 12 t, almost four
times larger than that of the C350. There are also several intangible
factors such as ease of SCR replacement in the field and slightly better
reliability of the larger device since it will run slightly cooler under
normal operating conditions.
21.3.4 Inverter SCR Selection
The current waveshapes of a 1 kHz sinewave inverter can be
seen in Figure 21.13. The SCR must block 900 volts.
I
h-400AM'"
---+
C\
• t
100,.SEC
----1000 pSEC--FIGURE 21.13
EXAMPLE SINEWAVE INVERTER WAVEFORMS
From Figure 21.8, I(RMs) = 89 A. The following 110 A(RMS)
SCR's have been chosen from Chapter 22 for closer scrutiny - C52,
C150, C154 and C158.
Since the C52 and C150 are phase control devices, the dynamic
stresses imposed by the circuit, such as turn-off time, dv/ dt and di/dt,
will cause these SCR's to malfunction. The upper blocking voltage of
the C154, moreover, is 600 volts, which is 300 volts short of the
required 900 volts.
Figure 3.19 in Chapter 3 shows the maximum, allowable sinewave
current pulses for the C158. At 1 kHz, 100 p,Sec pulse, the peak current
is 450 A. Since this device is also available in voltage grades up to
1200 volts, it is the natural choice for this application. The commutation circuit is designed to exceed its maximum turn-off time by a suitable safety margin.
613
SCR MANUAL
21.4 CHECK LIST
Measure and check the following against the component specifications.
Mu.
Laad
Min. Starting
Laad Load
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
Peak forward blocking voltage
Peak reverse voltage
Rate of change of turn-on current at operating
frequency
Forward current before tum-off
Average forward current
RMS forward current
Peak reverse current
Surge currents
Maximum gate voltage before triggering
Maximum gate reverse voltage before triggering
Peak gate triggering voltage
Peak gate triggering current
Peak gate power
Average gate power
Gate voltage rise time
No spurious signals on gates
Gate pulse width suitable for the circuit
No undesired saturation in magnetic core reactors
No undesired saturation in magnetic core transformers
Power supply impedance
No contact bounce effects from mechanical switches
Electrolytic capacitors checked for high AC cmrent
Case temperature
Operation satisfactory at maximum ambient
temperature
Operation satisfactory at minimum ambient
temperature
REFERENCES
1. "Take the Guesswork Out of Fuse Selection," F. B. Golden, The
Electronic Engineer, July 1969.
2. "Solid State Incandescent Lighting Control," R. W. Fox, Application
Note 200.53,* General Electric Company, Syracuse, N. Y.
*Refer to Chapter 22 for availability and ordering information.
614
GENERAL ELECTRIC THYRISTOR AND DIODE CONDENSED SPECIFICATIONS
22
GENERAL ELECTRIC THYRISTOR AND
DIODE CONDENSED SPECIFICATIONS
This chapter is primarily devoted to condensed specifications of
General Electric's thyristors, thyristor assemblies, trigger devices, and
diodes. These specifications are intended for reference only. For full
information the designer should rely on the complete specifications indicated for each type.
The Selector Guides for Phase Control and Inverter SCR's are
laid out to provide quick recognition of the four major selecting parameters - current, voltage, speed (for Inverter SCR's) and package. Other
important parameters such as surge current and dvldt capability are
also included. Comments highlighting unique characteristics, package
types, etc., are included to aid you in your SCR selection.
Unabridged specifications should be consulted for detail design
parameters.
The initial selection of an SCR starts with identifying the current
requirements, since this offers a measure of SCR pellet andlor package
size. SCR's are generally categorized by mayimum allowable RMS
current, IT(RMs)' The designer is cautioned that actual SCR current
capability is influenced by the
•
•
•
•
cooling system
switching frequency (more prevalent with inverter applications)
ambient temperature
coordination of SCR surge current capability with system current limiting (fusing)
It's prudent to check the detailed current rating information
provided on the full specification insuring that the SCR's maximum
current rating exceeds the worst case use conditions.
"Phase Control" is a term used to describe SCR's where fast turnoff time is not a prime requirement. The trade-offs in SCR design are
such that tum-off time has an unfavorable relationship to current and
voltage capability for any given junction size. Primary application for
a device with relatively slow tum-off are AC phase control- hence the
name "Phase Control." This type of device is also used for zero voltage
switching and select pulse applications.
Inverter SCR's are characterized for tum-off time (commutation
speed) capability and other speed characteristics. When designing for
speed, the parameter trade offs must be carefully weighed. Thus the
large matrix of speed, current and voltage capability for inverter SCR's.
As the name implies, major applications for these devices are DCIAC
inverters. Additionally, they are used in cycloconverters and other
pulse applications requiring high speed capability.
615
919
T '"
5
:e
t
C')
z
....
C>
...
f
G>
C')
C>
...
...'"
en
C>
v
:.;
o
v
iii
o
u
N
s:
(jj
0
."
'"
.;;
'" '"
C')
[
'"'"
Z
~
VOLTS
elor
C;}6
C>
I>
[
I
a; [
N
~
a
a
I
s:
... c
... ,..,"
"
en
[
C220/C222
'"
~
N
I
0 (
()
... "....
...
,.:z: z
'"
'"
~
,..,s:
"s:
.,'"c
..
o
0
o
[~~
-
I•• _C1111C)
-_IE
~+.
,T,
'fJIIIcI, ........ (,deIJ
.... rdHk1..........
. . . . . . . . . . . . . . . . . . . . ,....c-t)
-.
cao.U.1l.sa
CI.
lAG
U
U
U
25.0
'AG
OJ
10.0
@3SOC
••JOe
27
27
u
30
.. to
30
.. to
50
50
.. to
.. to
..to
.. to
'50
................. ,...
:=!crltICIlratHf.rlMof .......
. . . . . . . . . fJIV//MC)
I SPECIFICATION sam
NO.
_ _ III"C T, IITI'"
WI l WITI l III m nru.
50
150.20
~
IENIIIII. PI.__
150.21
150.22
150.35
SILlCGIE ~lEO WIIH PIIIlII IUS NlSlVAlEO PllUT.
FOR COST COISCIEIICE HUlME -.:ATlONS.
-1E-.-IIIL--PII-.PGS-L-"-.-mu-.-,u-um-IIL-amm--AT-,ONS.----------.J
3 _""mUIITII".
•3 PlCIIIIE
DIU.. CURIENT
FUTlllNI PIIIIIII IllS PllUT PASSIVATlO••
CO _ SCI
_TlONS.
150.36
ISO.30
180.19
1&0.27
~ T~ f1 ~ ~'"
---:--------------' 1:~'
i18
--
... .... ...... "".222
'"
...
...... ..- ..... a_ .... ....
"-I
,...
,...
..
1."
•., ....,
,.. .....
.
,.... ..... .....
.....
"
...
,
.. .. .. .... .. . ". ...'" "A..
,.. .. ,.•
..
_
.
"
... ,15'.. • ... ,.. ,.. ,....".. ". '""
.
. .. .. .. . .. .. ,
CIt
111'I711-n,- 1IH""7I
~,,-
~_~_'
_
160.23
It' "
,
JR.
k._l
~-
~•.
~ ~ ~~~O_'."1
I
GENERAL ELECTRIC THYRISTOR AND DIODE CONDENSED SPECIFICATIONS
PHASE CONTROL SCR's 25 TO 35 AMPERES
.. ...
".... ..
,
I~':"
SP£CIfICATIOfIS
I VOLT.IE RI.IE
COIIDUCTION
ITIUlSI
hlAwl
..... IIfIS ....... arnntlA)
_.
..... - . . . ......... c.nnt@1.
t:CIMIctl"(A)@To;t-C)
h~
,.
."'"
T,
'"
"'_(Al_
-
25.0
35.0
35.0
16.0
@65"C
@we
22.3
22.5
@700C
........... ··CfClI.n...,.,etitins....
..... I"t ....... '.5 ..
100
....ell.. .,.III. . . . . . . . . . . . . . . !"'C)
65 to
125
.. "'"
22.3
.
22.3
@we
300
320
"
65to
65 to
150
65 to
125
10
"
100
160.30
160.45
12'
..~I
...1_1
.60
,.
"'r..ime-ot-rlse of on. . . .
ClBTelltW,.SH)
''''
4010
125
OfF.STAlE
I"'"
I
MIn.crltlCIIlrate-of-rillofotr.slate
' ................. @ .....
20
T>p.
ratelllTJ(V/p.HC)
SPECIFICATlol SHEET 110.
160.22
160.20
-~--.-},A
IAIII lAm:mES.
THE ORICIIUL ER. FOR IDEm . _ E
UftlCArllllS. .1 REL mES ".IUILL
SIMRAI TO C35 ma
mEPT 15rC TJ an•.
HIIII 'l1li.1• . HIM PEIlfIIIWICE FOI DEMIIIDIIII
IlOomllL APPlICATIONS. HI m TYPES IVAUII.£.
t3
i~
100
1
160.451
g~
SIMRlR TO 1117.
619
SCR MANUAL
PHASE CONTROL SCR's 55 TO 200 AMPERES
...... . ..
HIM
JIIH
.,47
.......
,....
.t...".
CI,,'U
..- -,. ,...... - -,.. .. .. .... ..,. "
..... ...
....... _trdt._........... _
.....
'"
_FlCAT_
I WlLTIIE . .
e_
_ _ _ - - . . . c.nal (I)
h ....
h...,
h~"
,-
...
=-~~CUfHt.,
1IIL-.p ........ -..t81J0"'
-"nlll@Te
@93"C
7DD
70D
0"""
1IIn.cr!tlCllfItHf-rIMofol-stlteWtHlp,
... . - - ' ........... TJ (V/1lHC1
SPECI.ICATION SHIlT 10.
~__
.....C
"POSE IPfUCITIOlIS.
4010
"10
12&>c
3OTYP.
2DD
2OOnp•
3OTYP.
'"
170.17
170.18
170.19
~~
~
LOW PIIIIEIl lATE RElllIIEIIEm.
I.e TJ IAn.
fill MTIIIL ClllMCTION APPlICATIONS.
J-C
III. */IL mEWNI fGl IIDTOI CllllTllIIS & POIIEII SWPUIS•
... */11 lUI _
620
CIIIIEIIT _ _ Em.
'"
59
@'7'>C
12SOC
"''''
'".200
"40 to
EICIllEIIT */11& .1/11 UTIIIS.
PIUS PAl fill 9
......C
I ....
2000
"
"10
.000
-OIl
100
,.,
Ull.M
170.85
170.86
. "....:,-,-:-
l1li _ l I E lISSfS. UIW VlLTAlL 11811 C••INT,
WlTEI COIUD COII_TlOIl
622
CIa
~
....u •
1II'C
...
tcffJ
tcffJ
,~
~'
~
I
I
GENERAL ELECTRIC THYRISTOR AND DIODE. CONDENSED SPECIFICATIONS
22.2 INVERTER seR's
1400
-
1300
1200
r--
1100
I--
!-
1000
t--
!--
....
900
I
~
~
'"
...,~
700
J
I~
!-5
"
<;
I
"...I
">
~
~ 800
'"
....
"
~
...,
..
!-">
I
r-
!-
600
>--
r
10
~
1--
500
~
400 >--
:;;!--<;
J
.
300
,
I
200 I-100 I-0
r--
'-
:a
25
~t---
<;
!---
\l-
......I
...
~
!!!
\ l - "r--
-
-
...
~ ~
" rr-
r-
~~
-
35110
RMS
180 225235
CURRENT
-
370450550~ 700
AMPERES
SELECTOR GUIDE
INVERTER SCR's
623
SCR MANUAL
11.1
:IE
i=
I::
o
I
Z
II:
~
C!)
z
i
11.1
II:
U
~
11.1
:IE
i=
~
I
Z
II:
~
z
C!)
iii
C(
11.1
II:
U
11.1
o
Although each type of High Speed SCR is affected to different degrees by various conditions, the
directions of change ramaln the sama. Additionally, the ralative effect of each is roughly as shown
above for a "generalized" High Speed SCR.
Should you determine that your conditions are more severe than those for which the tum-off time
is specified, you ara Invited to contact General Electric to have turn-off time specified at the dlfferant
conditions.
624
GENERAL ELECTRIC THYRISTOR AND DIODE CONDENSED SPECIFICATIONS
INVERTER seRfs 25 TO 35 AMPERES
FOI COST SEflSlTIVl _
AmltATIDIS.
ClllllAClDlZID TO 25 1Hz.
. FOI fIllS( _ILUOI , ..." CIJTICAl ... _IC&TIOIIS.
CIIIIIAClDIZID TO 25 1Hz.
FOR ..._
, CRITICIL ... _TIOIIS.
III'TIMIZID III YOLTIIE , fUIII.eFF.
FOIIlftlATIOIIIIIII_DlIIIIL
HIIUII TO C13I; mIlE FOI
CHIllCTEllIlED TO 25 1Hz.
II£VEIIS(
YOLTAIE IlPElATIO..
0"1._ FII IIILllIE , lOIN Off.
_ T i l l ....
625
SCR MANUAL
INVERTER SCR's 100-380 AMPERES
--.. ...............
_
.mE
....
---.
e'c;=IPC.""...,W
....
.,• .......
-=-.I.-::==-
e..
IIIPLln....
III
DIFflUI
".
"
::.n.-_
u _ _ .................
~
@ . ., , . . . , . . . . .
ct... , .
III
"'
".
.-..
... .-..
..................
..
11:,,,'17
III
........ ........ ....
,- ,- ....,...
_lUrlllllS
l_r.. 1IME
h_
110114.'"
1200
105
m
os
1200
180
11•
100
100
140
1&00
&
e_//IDIC"-'IiIII
Cd6aI
_ _ Wil_
....................
"
"
20
30
"
35
20
1200
100
100
100
300
-....
"""no.
-,.
.........
140
110
120
125
1200
1000
10.500
"
"
25
lOll
100
20
20
30
."
".
III
DIFFUII"
,'"
".
21'
,lOlL,
...
320
275
110
215
115
170
,...
......
.
"'"
50,'"
20
"
000
lOll
lOll
.......
200
200
170.53
170.57
40 to 125"C
arrmrE
.lIt.altial,.....,.............
....... ......-. ....... Vmu
200
100
118.35
"....
@_.TJrr/~
_rlllllSlllTIIIL
~
_ _ TlE_IUllE.
_l1li II 1ft . . - Y . UIII SWIIC_ LISSlJ.
181 _
UlCUS,
camus & IIIT1III COIllROlS.
_III
.......
170.36
200
170.37
\1
170.37
"0'"
_
1
l i 5 em.
- . IIIH
1 1II11ME.
__
I Dl_
IMftJfYIIII
LeW _ _LIISIB.
=,&I.111...:::Ir.-=_
TIE AVAIUIU.
PUSS rill ...... If Cl"
_Hl_GfCIII.
U I I I _ I . IIISIS. FAST _ .
l1li ..TIDY IIIIW I IIIEIITIII . . . DInE.
PUSS HI _
626
e.
. ......
. . . . . . _ _ _ _ _ _ ..... c-c)
T,
..,.
180
"0
"'
..
III
10,500
@1_'........
"'.
"'
,-
....
........
,-
Gf ClI5.
v
GENERAL ELECTRIC THYRISTOR AND DIODE CONDENSED SPECIFICATIONS
INVERTER SCR's 550·625 AMPERES
,,-
hlllYSJ
~~e~.;.ft; :111""
II . . .
II .....
55'
• 5O
'70
@2IOD1ir
'"
130
....
Mil. III..... cre'l, .......,.Itlt. lurp
clI'lfttlAl
IIIH
.... '....."" ... 'I.
LS . .ICWSlc}
120,000
@200V/ltllc,...,l....
TJ
Crltlcalr.t.." ...ln" ........
c.rnntlAl.lIIlC)
J..ct.....llfltlnlt~r.'..... I"C)
I..
.... aUlui nbI...·rl..........
WlIIItIp .......IItiII .....hlly_
...... TJ(V/jtHC)
ECIFICAnON SIIED NO.
lOW ,wneNllII LO..... FAST TU••.oFF.
FOR BAnERY mlCU & INYERlER MOlOR
LOW swm:NINO LOSSU. FAST IU.......
FOR BAnnY mlCU & IIMItlEI MmR OlIVE.
IIiH VOLTAIE. FASTl1II......
FOR .DIIIM FR"UEIICY AWUCATIOIIII.
LOW VllLTAIt
fill
~M
FlllUEllCY oVI'LlCAllOJIS.
1l1li VOLTAIt
FOR LOW .....EIICY .vI'LlCAnolS.
-''''
550
55'
53.
.25
120
'"
'"
430
,......
5500
5500
120,000
120,000
15
30
......
53•
.55
225
37G
'"
....
130
..
15
11
100
100
..
35
.f·tUIl
,.1_
.11-,.
625
......
625
'50
265,000
175,000
..-....
35
25
30
50-"
'00
...
...
.......
......
170.42
170.44
170.44
'DO
110.7&
...
170.77
"'" 1
ncl.7'
I
aa
'. -
0
8
~.
~
~_-----.JQ
-HIIHIII.LJW-LOIIo_ _ _
FOI LOW FREtUEIICY _ICATIONS.
....
' .. UTE
120
170.42
a
8/I/Y(.
HIIII VllLTAIE.
FOI _IUM FREllllDlCY mLlCATIDIIS.
,--
'·1 liTE
40toltsOC
IIfF.$lAlEl
I"/fi
II
BOD·11OG
..
@UIOY/,IllIlc,..,,111II
dIIft
.."'., , ....
....
IE mE
CONSTRUCllON
st'ECIFlCATIONS
I VOLYAIE MillE
CONDUCTION
.
~
627
SCR MANUAL
INVERTER SCR's 625·700 AMPERES
...,
lIE
mE
: COIISTIIUCTION
SPECifiCATIONS
t V81.TABE WIlE
COIIIIUCTION
"-
....
........
........
........
...... ,.l1li ....
m~~...
SOD·'.
,-
.so
CIIrnnt(l)
.SO
erell, n..oftIIItfth" ..rp
1Iu..lttt.minl'"
U . .t(l'lH)
~
'"
'50
'50'
230,000
@
T,
".
"
1l1li·"
.
300
'90
'50
,30
'500
230,000
65,000
30
"
...
2DlNh.sec ,"",led
Crltilllllirabl..r-ri. . . . .___
CIrJ'eIItWp..1f/
1uIctI1II"'-.ti.. tMp~....,.{"C)
OFF"'ll£
...
......
"0
SPECIFICATION SlIm NO.
170.45
17".45
.70.ao
IIBN_.
III LOW nnulllCY UftACATIOIIS.
a
FUll CON_NT CllAUCTERIlATION orTlMIZED
FlII 11Hz SWITCHII..
MEDIUM
V81.TAlt
FASTAPPlICATIOII$.
_.
FO.
_
_ ICY
LOW SWlTCHI. UISSES. FAST TUR...".
III BATTERY III1IICU l INVERTER lIMO. HIYE.
LOWV8I.TAIE.
.1IM II'EED
FOIIERY Hili._
_ • InERTElllIMOI CIIIIIIIIIIS.
628
em
AIIIPL'
",".
1m
,.....
...
......'"...
...,
8000
250.000
250,000
...
."
5SO
100
"
40 to 12SbC
.......
MIIII.crltlalnte-et-t'lslef.ItIta
" ............blll.ratedV....
,--
""la'll'.
15
@ .... TJ (v/,d.,
,",..
I
...'"
-,
",
Tlrll4ltl_ttntell'lOlllp&
~ '1\eY/:'~ ~r)
@,oav/,llSler..",""
"'"
f_,IATE
MM. . . . . . ClMUctlM 11n1lSllJll'-1
@Tc=IrC.IID'" ..., (I)
@1.111:
.
....
...
1I~~11Ii
a
200
'70.42
170.42
~
~
([;jJ.
U
J
I
GENERAL ELECTRIC THYRISTOR AND DIODE CONDENSED SPECIFICATIONS
22.3 TRIACS 6 AMPERES
Jf.
MTS
Gate
MT1
GE 'hili
Vol.... Rang'
Current Rating (Amps RMS)
ITSII, Mu Peak On. crcliA)
Nn. R~. Surll CurrHt A
~RII' alaclllng cijrrent
25°C, Mu (mA
dv/dt static, @ 1DOoC
Rated Y
Gate Open
Typical I /tIle)
n/dt Commutatlll, @ 7SoC
Rated VDRII and IT; Gate
OPH Min (y//tlle)
Iv':"
FIRING
laT Mu DC Gate Trlaer
Current @ 12 V, 25°C (mAl**
VaT Max DC Gate tria.
VDI....-@·12V 25°C(V)**
PecllU:I Type
Specllicatl.. Slleet No.
SC240
SC240*2
SC241
SC241
200·500
6
200·500
6
200·500
6
200-500
6
80
80
80
80
.1
.1
.1
.1
50
50
50
50
4
4
4
4
50
50
50
50
2.5
2.5
2.5
2.5
Stud
...
Isolated Stud
175.16
Press Fit
Power Pac
•
175.15
*Voltage Grade
**MTo-I- Gate-/MTo-I- GateMT..-Gate-
PaW,r Pac
....... Flt
Isolated Stud
Stud
629
SCR MANUAL
10 Ampere Triacs
liE tiDe
SC245
200·500
10
VDltap Ranp
Current RatiiiiliiiiiiS RMSI
ITS", ~:.' Peak One CJcI'j(
NDn. Re • Su...e CUrrent ( )
::..., IIHId... CUrrent
25· C. Max (mA)
rlYl lit static, @ 1OO·C
=
-r:
lcalrl'vRJ'uII~
I Ie apR
rlY/llt Commutatlng, @ 75·C
d VD·~l:s~ ~f· lIate
OnMlnlec
FIRINII
IGT Max DC lIa1e Trlaer
Current 12 V, 25·C (mA)**
VGT Max DC lIate TrlUlr
VoIta,II.' 12 V, 25·C (VI.*
ii
Paella.. TVDe
SpecificatiDn Sheet ND.
10
SC248
200·500
10
100
100
100
100
.1
.1
.1
.1
50
50
50
50
4
4
4
4
50
50
50
50
2.5
2.5
2.5
2.5
Stud
4
SC245*2
200~500
Isolated Stud
175.11
,.
Press Fit
SC14.
200-500
10
Power Pac
175.15
*Voltage Grade
**MT"fMT"f-
Gate-!-
Ga1~
MT_Gat~
....erPac
630
Press Fit
Stud
Isolated Stud
GENERAL ElECTRIC THYRISTOR AND DIODE CONDENSED SPECIFICATIONS
15 Ampere Triacs
GE Type
Ya/tage Range
Currant Ratinl (Amps RMI)
ITSM, Max.Peak One C~I(k)
Non_ Rep. SUl'le Current A
~R'" Blocldnl Current
25° C Max (mA)
dvl dt Static, @ 100°C
Rated VDRM Gate Open
Typical (v/J£Sec)
dv/dt Commutatinl, @ 75°C
Rated VDRM and IT. Gate
lipen Min (VI J£Sec)
SC250
200-S00
15
SC25O*2
200-SOD
15
SC251
200-500
15
100
100
100
.1
.1
.1
50
50
50
4
4
4
FIRING
IGT Max ~C Gate TrilPr
CUrrent
12 Y. 25°C (mAl*"
VGT Max DC Gate TrilPf
Valtage @ 12 V 25°C (V)....
SO
50
50
2.5
2.5
2.5
PackapType
stud
..
Speclficatlan Sbeet Na.
IsolatIKI Stud
175.18
..
Press Fit
*Voltage Grade
**MT.-+- Gate+
MT.-+- GateMT..- Gate-
Stud
lsoleted Stud
.....ss Fit
631
SCR MANUAL
25 Ampere Triacs
GE Type
Yoltage Range
Current Rating (Amps RMS)
ITSM, Max Peak One Cycle,
Non. Rep. Surge Current (A)
IDRM, Blocking Current
@ 25·' C, Max (rnA)
dvl dt StatiC, @ 100·C
Rated YDRM Gate Open
Typical (VIILsec)
dv/dt Commutating, @ 75°C
Rated YDR(V) and IT. Gate
Open Min I ILsec)
FIRING
IGT Max DC Gate Trigger
Current @ 12 Y. 25°C (mA)**
YGT Max~C Gate Trigger
Yoltage
12 V. 25°C (v)**
Package Type
Specification Sheet NO.
SC60
200·500
25
SC60*2
200·500
25
SC61
200·500
25
250
250
250
.5
.5
.5
100
100
100
5
5
5
50
50
50
2.5
2.5
2.5
Stud
175.28
Isolated Stud
175.28
*Voltage Grade
**MT2"\- Gate+
MT2"\- GateMT2- Gate-
Stud
632
Isolated Stud
Press Fit
Press Fit
175.28
GENERAL ELECTRIC THYRISTOR AND DIODE CONDENSED SPECIFICATIONS
22.4 TRIAC TRIGGER DEVICES
TRIAC TRIGGER DEVICES
Because of its bilateral nature, the Triac requires 118 own unique trigger element, distinct from SeR's. GE offers a
full line of suitable Triac triggers. The total power system Is easily assembled consisting of a suitable sansor, Ie.
Triac and a few other passive components. The Triac may also be triggered by an SCR trigger element by using
a pulse transformer.
APPLICATIONS
UIT
DEVICE
PUT
S8S
2N49!11-93
ST2
ST4
USE
HEAT CONTROL
LIGHT CONTROL
MOTOR SP£ED CONTROL
POWER REGULATION
SOLID STATE
CONTRACTORS &
RElAIS
1=IXCEIIfJII'
F,.,FAIR
P=POOR
N : NOT APPLICA81.E
CIlllVEllTIOIlAL UNUUNCTION (OJT) AND PROGRAMIIABLE UNUUIICTIOII 1IANSISIORS (PUT)
• Unilateral triggers requiring pulse transformers.
• See unljunctions.. switches, & triggers selector guide. Specification No. SO.53
DIAC (SI2)
• Dlffuaed silicon bl-dlrectlonal trigger diode.
• SpecfficaHon sheet No. 175.30
SIUCON ASSYIIMETRICJL SWITCH ($14)
• Assymmetrlcat trigger device for hysteresis free lamp dimming circuits.
• SpecHication sheet No. 175.32
SIUCON BllATEIIAL SWITCH (SIS)
• Low voltage triec trigger, two silicon unilateral switches connected In Inverse parallel.
• SpeclficaUon sheet No. 65.25
633
SCR MANUAL
UNIJUNCTIONS. TRIGGERS AND SWITCHES
Since the introduction of the commercial silicon unijunction transistor in 1956, General Electric has continued 1111'.
E'
>...... SIIIde
J ~::.7
.....
<1"
,
::
.....
j~~===---------~~---+--~----~------~~----~------~-----7~
634
GENERAL ELECTRIC THYRISTOR AND DIODE CONDENSED SPECIFICATIONS
CONVENTIONAL UNIJUNCTIONS
General Electric produces a very broad line of standard UJT's. The T0-5 ceramic disc bar structure device has been the workhorse of
the unijunction industry for over 10 years. MIL versions are available on the 2N489-494 series. EquivaJent types are available in TO-18
packallfi where small Size is required.
The cube structure TO-18 series offers excellent value for those requiring proved, low cost units.
sc._
""lHIIIns
11.,11-'
TI..rs
- ---
-...
...... """=1'1
I."".
..........
.....
.......
--_._..
_.
-,
" .....
----............._.
......
---
....,a.......
f ...
SawtIoIIIII.mtan
SllbltYlltap_iII
-
g
6.2-9.1
.Ii
~
4.7-8.8
12
•
•
.56-.68
2II,ma
......
...."
2It2417A
....'71
....11
-,-,...
.............
'"",.
_II
'"",.
;" ....,..
""""
.I
,
Z;
~
.-
31
~I ......
......
=M
·
·•
·•..
·••
·
·•• .....
·••• ....
•
G.2
0.'
.02
2
G.2
0.'
.62·.75
12
12
0.'
.02
25
4.7.9.1
.47-.62
4.0-12.8
4.1-8.8
.51·.62
6.2·9.1
.51-.&2
4.7-6.1
.56-.&8
&.2.9.1
.56-.&8
12
12
0.2
.M
12
0.'
2
12
30
30
30
4.7-9.1
.47_.&2
4.7-9.1
.56-.75
4.7-9.1
30
30
30
30
.12 Typll:l'
~
~~~~~~~ It~~~0: :c~~manded
"B" versions in addililln 10 sCRtriuerinl
.
luaranleHlowerlEolndl,forlonltiminl
perlDCIs wlth._ller ~Ipaeitor.
60.10
110.10
----so.u
60.10
•
8o.tO
~
Industrialtypl$_
".53
General jlUrpOle-low COlt.
30
0.'
0.'
.M
30
30
30
30
30
60.18
¥~r~i~~: It~~·~1doJ:, ~::meftded
"8" ..rsionsinaddit'ontoSCRb-igerl",
IUlrantees 'ower 'mand 'pfor '0lIl tim.",
pwiods with a smallwl:lpacltur.
30
~
8
60.10
80.10
30
30
..
30
"
No.
30
...,.,
..
"
....
..60.10
30
30
30
.&1-.82
4.7-9.1
4.7· 9.1"
•
-..
'"
30
G.2
12
"::'
30
0.'
0.'
-~
30
0.'
2M.....
.....
12
12
..
..
..
.....
..
....
TJ_:IS"C
.02
.12-.75
""'"
.......
_II
12
12
12
12
2112421.
-,.
......
...
...................
, .. "-
·
·•
·•
·•
"
·
"•
""
""·
•
"
""· ••
·
"• ·
""• ""
12
12
'0..71
211'011
2111..,.
M••
12
..-
i!
.......
,..r.., • •
..."',. :::.
..., ....
....,
@'-=1"
4.7·6.'
4.7-6.8
~
/--- ..- --
30
JO.llvenlOlll of2'U871 industrial series.
Gener,'purpose.
60.62
Far'CI/lItimillliperl.....ndtriuerilllihilh
ClntntSCtt's.
GeIIera'purpose.
80.12
Glnllra'p~'owcost.
60:13
Forl.5vottappllc:atiOlll.
• JAN & JMTX tJpes _11Ib ..
'''''=I.5V'
635
SCR MANUAL
PROGRAMMABLE UNIJUNCTIONS (PUT-D13T SERIES)
The 2N6028 is specificaHy characterized for I~ interval timers and other applications requiring low leakap and
~~kir...::
2N6027 has been c raeterized for general use where the low peak point current of the
.:e"riW!iThe
_-
_.. .. ......- .... ... ....
_.....
.:
--.., ..........-..., -,, ="" ·"*1'''_
...... -..... =
, ,.
•
• .... .'.'
..
•• •
•
"
--.... --,
...,
fm
III
1'1
,SO
t:'
1:: ::-
&
L
Yo
.~''''
(MII.I
10
'50
.15
...
10."
10."
COMPLEMENTARY UNIJUNCTIONS (D5K SERIES)
~
..
...--
..- -..
---- -
....
• ... tol.1 ....
5.5-1.2
~15
.,
-..
.=.
.... '=" ~
.... "'" ......, -...
71.
......, •• • ""•• ......
.~
&
...z..
:=
Yo
OIL
__ ",,+11PC
1'1
.58..62
10
-55to+J50
55to+~OO
15
..•
-..
2A
_.
...
I
10.15
I
10.18
SILICON ASYMMETRICAL SWITCH (SAS)
-
.".,
636
.
....a::
1'1
I ".II'
..
':.'!:'"
-=-1,;II
.
,'='
......., ....
1110111
'ft".:1'
1'1
10
..•
v.,
,-
'ft".:1'
MI ••
1'1
1".11"
10
....
'=-
_::'.'.2I"C
.03'1fo1"C
-..
_1M
...
.....
GENERAL ELECTRIC THYRISTOR AND DIODE CONDENSED SPECIFICATIONS
SILICON UNILATERAL AND BILATERAL SWITCHES (SUS, SBS)
The General Electric SUS is a silicon, planar monolithic integrated circuit having thyristor electrical characteristics closely approximating those of an "ideal" four-layer diode. The device is designed to switch at 8 volts with a typical temperature coefficient of
0.02%/oC. A gate lead is provided to eliminate rate effect. obtain triggering at lower voltages. and to obtain transient-free waveforms.
The SSS is a bilateral version of the forward characteristics of the SUS. It provides excellently matched characteristics in both directions with the same low temperature coefficient.
..
''''
.
......
.... ,-,
...."AI
,~
,.
C... lill....
,~
V,ttql
~,
.
,-- ,_.... -...
-",. ,... '-
,.
........
,....,
.<
@IWC,
lOll' •
'%
"
Dltt_1II
~,
S:I::-"
(~
.00
30
"
I')
",I)
(W)
to
_M
!i;g
to
....,
!
200M'
=.02
I.'
I.'
"'.05
I.'
.....'M .....,. ,
:= ,oA,
..U
• Measurtd@I25"C.
V.
--
'"
G
~
~
65.26
..,
U
..."'"""
-.::~
6!.2!i
-----as:25
-----as:25
~
,.,
~
.20
65.30
High tripering sensitivity. 4 lead capabiiity
for multiple load or dv/dt suppression.
. ..
_.-
':':'
,.,
SILICON CONTROL SWITCHES (SCS)
""" '"
(Ill)
"'"
''''
""
....n
~
Cf)
==
I.'
,",Nt
1!
..
,.
...-
"
~~ ~ =i?
tll 5
CMtl......
DCf_1NI
''''
200
- -..
,..
....... .......
-....
...
"'
bristles _Idle$:
,.
_
..
-... ...
-......
,_.
..
1'..':
.... .....~!;
, , '" '" 0..=
,-,
'''"'''
,oA, ,..,
'';: ...,
,AI
" "'" ""'=
'"
""
.""
_ _...1
c.t1l1Dd'
@tOO .... K
101l~!
..
tOl!!
(MA,
1&01=
..,A
~,
4.0t
@'IAIl= .....
........M="'.
'5''''
=~!.
lOA'
1&0<=1011.
'"
'......
"
65.16
10
65.11
.4to.65
65.11
175
t ....ured In speei.1 tnt
N••
65.11
circuit (See speeification sheet).
ADDITIONAL REFERENCE PUBLlCAnONS
ORDER· BY PUBLICATION NUMBER
90.10
90.12
Ttle Unijunelion Transistor Clulr.cte.ristic:s
and Applications
l/n{juJtction Tetnpefat1lTe Compensation
90.19 Unijum:tion Frequenq On.ider
90.10 The OI3T-A "OIOIII....le Unijuncticm
to.72
c~
Unijllllction Transistors
Tral'lSi.tOf
637
SCR MANUAl
22.5 DETECTORS
,_eta
to_
......11_. hu a "'" n.. 01 II.IId _
0,...I00I'''''''' _ a n d 1I1_ln_ngYGUch_
your fUnatlon wIIfI mulmum IIfIIotIvenuI and minimum COIL
ThII
Selection
Glu"
oontalftl
Informlllon
on
prodUOII which hIVe btIn tormlilly IntroduHd prior to III prllll1ng. General EleCtrIc
wlllbapubl _ _ _. _ _ _
hili oontlnull'll dlvelOplllllll PftlClrarft and wDl be lnlroduclng other productl throughout the year. InlOrmation on th... products
PACWES
ThI dtvl08 mUlt have. lunkJlently .... rile and fall time to ....... ligna' In the cael of • moving aperture or a Ilgnal from •
pulRdaource.
SEIISIlIYITY
.....1I1Y1t)' Ie Important to gl.,.ra. a .Jgnal .afflclant to drive • loglo gate or"oth.r foIlowtng funotloM.
SATUUnOll V8LTAIE
The "ration II vet')' Important when the device 'I being UI8d to drive • logic aate dlNOtIy.
1YPES
................ II1II.11I11III
u_
UK
Lim
Ll4I
L14I-..
u_
LI.
'"
WIt
638
---..•.
•
....-
.....
=••
••
••
•••
h
••
,.
..
II
I
II
OA
-""'b,
•
.
10
I
101
II
"""""':.,
...
.
~
OA
1Wtk.... (lGDIIIQ
.7
'ID
........ (l7l.a)
IWIICIIHU.tAIIII)
--.. ..........-...
SpecllIUtI..
A
A
LI
.7
A
...
,~
La
....,
.....
.....
.....
lUI
.ID.IO
GENERAL ELECTRIC THYRISTOR AND DIODE CONDENSED SPECIFICATIONS
PHOTON COUPLED ISOLATORS
ISOLAlIIlt
The device must have sufficient isolatIOn to prevent destruction Of the devk:e.
SPEED
The device must be fast enough to pass the ,ignal without distortion.
TRANSfER IAnlt
Transfer ratio Is Important In thalli can save you an addlUonal stage of ampllncatlon.
PACKAGES
1/
/
~
~w.
•
. -........
TYPES
........d.d _kill iIItn. ia Hid
.......
.....
.......
H10'
• 101
• 1..
55.63
3.(11_
PC.,...
PC 4-..
ARRAYS
CENTER TO CENTER SPACUIG
Element must have proper spacing to be"compatible with format to be read.
SPEED
Element must be faat enough
to sea the moving aperture.
MATCHING FACTOR
Moat data formats (I.e. punched papartape. and punched carda) transmh some light when a hole is not preS8nl, hence we have
two signal levels (an on--signal, and an off-lSignal). II Is Important to determine both tevels. Once these levels are measured, the
required matching factor for the devk:es In an array can be eat. "to Inaure an on-slgnal from the least aensitfva device In the
on-atate, and an ofl-signal from the most sensitive
device
In the otf..Itam.
SEIISITIVtn
It is desirable
_.
to generate a sufficlant signal to drive a logic gate or following functions directly.
SATUlAnDN VOLTAGE
It is necessary that the phato-translstor have a low saturation voltape when It is being used to drive logic directly.
TYPES
''1IH(.d)
t.
J3DO
9.1OD'"
31
3
50
1.0
JJCWIA
!I
.lOD'"
S
:I
15
1.0
SpecIllI---PJuu CMIIct JOUr
IedrGnIc CMIpaatnts SlIts 0fIic, willi JOUr requlnllllllts eft all speclllllnQl,.
local"
.......
y=r
A
A
.....
639
SCR MANUAl
VISIBLE SOLID STATE LAMPS & DISPLAYS (EMITTERS)
·PEAI EIISSIOI
To Insure that the film lin tum ......... appIiCaIIons) or the eye' lin visual applications) are senalti"" to the emitted wavelength.
VISIBLE OUTM
Mlillcandelia II 81111111Un!1 ofth8Uma rate of flow, of visible radiation, out of ttl. device.
Package is Important fOr proper center to center spacing In linear and matrix arrays.
Package size is aIIIo Important since the apparent size of the light source is the diameter of the package.
sau12
sal.
saul
"13
_.
_.......- ==
....
U
UI
....
.....
....moo
. . . . . . . . . . . . @'
~
I
VISIBLE
TYPES
__
_ il"'ld
.....
.....
"'\'P
_.
.125'
.'17"
s.p.,."sefor
......
.230"
LOnprBlrreliforPllnl'
nr
.230"
-......
Llnelr MId Mltril Arrays
Mountinll(Panel Mount
Clip Available,
1.0rdll@IIIODA
SSL14D
SSUO>
640
=.=-:
....
_
DISPlAYS
....,.........
a.r:-
SpedI.=-
~--------~Am=--------2m~------~--------~.~~----------~~~
Am
~
~
GENERAL ELECTRIC THYRISTOR AND DIODE CONDENSED SPECIFICATIONS
INFRARED SOLID STATE LAMPS (EMITTERS)
POWER OUTPUT
Infm.recr Solid State Lampe ara used to Irradiate silicon photo-detectors. The output from the detector Is. proportional to the
radiant flux density incident upon its active area, hence power output Is the most Important parameter of the Infrared SSl.
PACKAGE
The package must have a small diameter for proper center to center spacing In arrays.
/
Tha speed of the device is important In pulsed operation, primarily lor signal coupling over large distances. In most applications
the speed of the detector will be the limiting parameter rather than that of the SSl.
INFRARED TYPES
-- -
R..._lOd.' slacklll_ln bold
,....
ssm
SSUlS
.....
ssw
.....
SSIA
SSLS4
SSI.5C
SSL...
SSl35
SSL551.
u.s
u
...•..
..,......
I.'
1~
•..
5.5
.oW.
.097'"
.""
......
....
....
."".
.2311"
.230.
.
.."".
.230",
Odllll
•7
.7
.'05
.
.010
.010
.3
3
3
.5
.3
..
.........
,......
...
..010.7
AlG
.,
.,.,
2
.5
.2
641
SCR MANUAL
22.6 SILICON RECTIFIERS
THE INDUSTRrs BROADEST UNE OF POWER REcnFlERS-.250 TO 1500 AMPERES, UP TO 3D VOLTS
• CURRENT/VOLTAGE RATINGS
• IUGH-SPEED FAST RECOVERY
• PACKAGiNG
• iRAIiSiENT SELF·PROTECTiON
• MOUNnNG AND COOLING
• GENERAL PURPOSE
PACKAGES
/
e':;:.
109.1
Ii: 1/
/
~
1"~~~
/
...........
,"
l
~ ~
I //
642
./
GENERAL ELECTRIC THYRISTOR AND DIODE CONDENSED SPECIFICATIONS
SILICON RECTIFIERS .25 TO 3 AMPERES
J£DEC
BE TYPE
SPECIFICATIONS
1.wlllVI
(~
M"'10·15
114PD
11141112
.,41·'
'"5124·27
GER4001.7 A1t4A-M
"5A·N
AI1D"
A15F
AllSF
.OS
@T"{,,C)
V~"'I .....J
-
lM50S'-621N4Z45-41
01230
"
"
50
"'r. repetltl" peak reverse wltqe
(¥)
.
,
...'"
......
...
55
G£14001
AI4F
GER40()2
OT2301
"SA
A115A
"'8
"1158
AI..
IUSC
oT230e
lMSO§
OT2308
lN424S'
GEMOO3
1"5624
OT230H
A114C
AI..
G{l4004
.,,,
A1150
1"5061
G£R4OO5
AI5M
'115M
1"5062
G£R4006
IN4246,
A14E
..
MPRIO
"
IN5627
GER4007
A14PD
MPRlS
hVI_1
MII.pta"enetrde,nln-rlClIITlllturr;e
wrre.t(lOlIJsluw••e,lphase
"
aplral:lon)@IIH.ratttlloaliconditIGns(A)
..... II.....petltifllfer •• 3mscc.(A2$ee)
T,
T.,
,~
•.
o,eratinrjllnctilntemperatureranp("C)
-6510
sterq;etl"plr.lture nIl' ("C)
-6510
Mil. peak f _ _ d wltlp drop@
ratelllFr""1(1 phaseo,lratlanl
'"
'00
40
SPECIFICATION SHEET NO.
"
--6510
175
-85"
ISO
-6510
150
6510
175'
6510
'51'
-6510
-6510
175
'"
Milt. r"erse reClP'err time (/lUt)
PleDGE OUTLINE NO.
40
3.'
38
130.53
175
1.1
'"
'"
30
16,
-6510
'00
6510
175
-6510
6510
-6510
1.2@
+·55c C
1.1
6510
'"
'"
'"
-6510
175
6510
-6510
-6510
'00
175
175
"...65""
175
I.'
0.'
n'
"
n,
130.55
130.69
130.55
130.56
'"
13Q.6fi
119
1192
1192
130.58
119.2
130.67
NOTE:
1 Averal' forwanl current 1 amp. @ r .. ::::90Ct. Junction, operating and stMlle temperature ranle -65 to +165OC.
° JAN 10 JANTX types available
643
SCR.MANUAL
. SIUCON RECTIFIERS 5 TO 1'2 AMPERES
,.,112·1'
lEBEC
tll1Mtl4A,
1II1II7" ,. . . .
.
lEma
SPlCm~~._.
IFllfII"I
.....'" ,.,.utift,....
@Tc-(-C)
W... f....!
5
'50
'50
10.
IN1341A
IN3879
,-
......... ,-
,
1111'....' _
12
J2
'50
'00
...,.
12
'35
,IU'I-,.
At . . . .
..
J2
r.nru " ..... 00
',lN1613
INI200A
1"l342&
......
INt615·
IHleUi"
1"3890"
lNI201A
IN1343A
IN181,,"
IN3191"
IN1344A
IN3IIIl"
UU202A"
lIU345A
lN3...
INJ203A
1~2
INI346A
lN3183'"
....
....
INI204A"
IN3893"
. ""
INI341A
INI205A
AI29E
INl3UA
IN1201SA'"
.,,121M
IN3117CIA
...
1"'1_1
''''
,o.
ere". _...-.1It
MIl. . .II ....
sarp cwrat
(IOlIIsIII................."ltI..)@IIID.
ratttllullc.llditillls{l)
.
.... _'""'"1ft
fw 1.0ms8 (AIsItC.)
T,
Opem:InIIII!lCtI.1II.......... rMl' C-Cl
T••
S1trapt-.lrlt.'nlnl.("Cj
.."
"-
M... till..... rnlstaDn, jl!nctiDII·to-use ("C/Wl
.'1. pill,..." YO""IP @mIIIIFr...fl
(1p1\Ue .....tilII11YI
@Tc-t-CI
.... Nnrn
rllCfl'_, tllll ( _ I
PACIIAIE OUnll1E NO.
'SPECIflCATION SHnTNO.
-JAN, &JANTXtJllhnllillble
644
"
IN3671&
lOG
'50
6510
+'90
8510
+'"
7.0
u
IN3672A
IN3613A"
IN4510
IN5331
1"4511
.50
65 to
"" ..
+'"
6510
+200
u
+'"
85to
"
6510
+150
+200
6510
+115
4.25
2:$')
..,
"
25
'20
'20
140.15
140.10
...
1N3919
IN""
'50
25
"
..OF
1"1199A
.20
140.12
..
" .. . ..
'35 ..
2<0
87
6510
+200
65 to
65 to
+."
6510
+'"
+175
".
2S
".
140.20
6510
+115
+12'
65 to
+115
+175
+125
+200
U
15
40 to
'00
.20
140.22
'20
140.23
.20
".".
6510
GENERAL ELECTRIC THYRISTOR AND DIODE CONDENSED SPECIFICATIONS
SILICON RECTIFIERS 20 TO 40 AMPERES
-- .....--..
_.... ........
1111U4D
1 . _. . . ,illtllA-llA 111:'_ IIIJ11.U ,'",. . . . . ,......., ' . . . .'3
'O5'"
IE mE
' -" -.
,..[1 . . . . . . . . . . . . (1)
.
10
..
,.....
111183
IN2155
111124!18"
,
,......
...
INIlIS
'"
IN2157
I ...
...
1"3lln
MOO
IN1187A
INl912*
'.3211
IN1l811A
113913·
'fll3212
Mo,
.."
1"11911A
IN3214
A40M
""
...,
..., "
...
A44M
A139M
1N3165
IN3766
...
-fA)
.... 11t ...... (JI-.H-
::::--..
.............
ttn .....J_ec..) . . . .
-~
~
........... ntliIbne.
~1W)
A139N
., ..
65to
--6510
+175
+200
+2DD
+175
65to
.+175
"0'
I.'
U
65 to
+200
IA
6510
65 to
_6Sto
.".,.
'200
'10'
=r~v~
..........
.,1')
-
I ..
150
140.28
1-
NCUlfaunllElIO.
SPRmcanllllSllElTND.
'"
'"
140.40
...
A139pa
'N453D
5llD
65 to
+175
MH.renrs• . . . . . . , _
A'"
... ...
115332
~---(II
III .......... ' .........•
au.J ..... nhlllIId
@Te=r-q
I.
.
113911"
MOE
__ ""_CICI&._ '"
.-....ta-a..
"
1I13!I01
IN3213
112160
I,
1:.~~~8
INJ20!t
IN2159
OlD
.
IN3909·
IIU18SA
INlI86*
INIl95A
1"3899
JNl184A
INl196A
""'~
INlI83A
.,..
.. .. ,..
"
,
• 'c-I"CI
~--.,........
,....... ,.3.,:,4
65 to
6510
_6510
+1"
65 to
+1 ..
+'"
I.,
...
'23
>2'
I .....
140.47
""
I..
6510
+115
-"
.
+200
_6SID
-I-J1S
,6M"
I.'
typic,'
1.35
typical
25
&'>'0
'"115
fi~
to
""
.......
....
+125
1-~7~_~
~i~1
1.0
Typlc.1
1.3'
1.15
140.33
140.26
25
...
'23
125
140.48
140.31
140.32
'"
• JAN • JMTI t,..es ,ni.. bJe
645
SCR MANUAl
SILICON RECTIFIERS 108-275 AMPERES
I
,,,....
J£HC TYPE
:1I111'E
.
.~----
SPECIFICA'IGIS
......... '-rilurr_ (I
In,I!A"
8
v...!'....
_
...
,-,. ,........
.-se .......
.01
.60
'"
Te=f"C)
Ihtr.,...... ..... _uYdlpIll
51
.N32IO
100
1N3281
..
,
A9OA.IN3735
IN3262
...
o.
A7CB,lI'fl289
"'"
A90D,lN3138
IN....
....
...,
'70S
A7ON,1N3294
..""
o\tOE,IM31!9
''''70
JN32l1
'90'
"lOT.
IN3272
"70P,IN3295
IN3213
A1OM,1"3293
ftOO
.."
"'.
113267
0\70£,113292
,...
""
A90C.IN3137
"700,1"3291
...
,.
MOB,IH3736
...
1N3265
A7OC,113290
SOl
.."'.,,'""'''..
-..
.. -
AtOM, IN3140
A90N,IN3741
........
A96'
A90P,lN3742
....
"'.
-..
'"
"".....
...
,_
215
.-.
._,
.-......
......
.-......
"'"
.....
.....
.....7
.-
...........,
"'"
......
".....,
'040"
...
"'IP
,
A70Pe,I"3298
. . . PC
.... PO
......
A291P£"
IfWl''''
,.
"
,'.
.,.
..
::~I-:':~;::'~~~""'"
rllellI_CIIIIfitiIIIi(A)
Satrap tIInfer*'e
n.,.
10,000
c-c,
fOCI
MIl. tI!emII,.......... jImcIi. . . . . .
P ...... .,.miIa)
"'<
'001
16,000
".,
+'01
+,..
+""
-5510
+I"
cec/WI
1II1· . . . . . . . . . . . . . . . . . . . . . . . . . . . .1M1
US
.,
'.6
."
3300
..., .....,., .....,.,
43.000
+200
+.25
+125
+""
+."
+.25
.11
.11
•18
1.3
....
'21
1.25
-40.,
• 30
SPECIFICATIIII SHEEr 110.
127
145.15
,.....
....
84.010
....,.....,...... curp (.Ie)
PACIIAR OmllE III.
646
..40',"
.... n.....epetitiq . . u.sec.(l'Jec)
CIIIraIiIII ..IICtiIn .......... ' . .
.600
......
121
......
43.'"
25
••
"IOU
•45.n
....
...
84_
'00.000
_to
+201
_10
+2110
."
..
'29
......
-.,
....."
+190
+190
• .35
.20
.21
......
I
GENERAL ELECTRIC THYRISTOR AND DIODE CONDENSED SPECIFICATIONS
SILICON RECTIFIERS 400·1500 AMPERES
...
IllI1I'£
_RATIOIS
~
.......
..... - . . . . . . .".11 .............1(1)
-...
. ...,..
.-
'40
.1c_("C)
. . . . . . . . . . . . . . . .11: . . . . . . . . . . .
........
00
'II
A390A
ao
Am.
-......
ltO
"II
,ao
.......
,
A310C
...50
A390D
""50
"'"
.....
.m.
"""
Al90T
"....
"05£
A295M
A295 •
"""
....
,.
...
""P8
ASO,,"
A295I'C
A500PC
A295PD
""""
.5OOP£
""'"
ASOOPS
""'"
A295PM
""PS
A295PN
T,
-..a.........nbn,.... C"CI
T..
- . . . ............ .-eJ
""
::--...:*.::=,....... ..,.,.....IFI""I
TN
.... _ _ IRIIItII•• ~.-e.nn
.'c-t-C)
-II.
_nilllIIm 10. .
.'708
......
......
AS...
AS'"
A510M
"5709
'IS'"
.......
.......
---
A540PE
A540"
""'PO
A500U
ASOOLB
ASOOLC
.....,
A500LD
.5OOlM
.......
e:.J!.~=r-~T=fir U . , " (Allee)
....
A540L
--....
....
--
..........._
..
...or
ASOOPA
....
OM_
'000
AS....
""PA
A29SPE
,ltO
......,
.....
"""
......
.....
Algor
....P,
-<
""'"
.....
10,000
15,000
14,000
200,000
270,000
400,~
920,000
-40TU
-40 TO
-<0 TO
_TO
-40 TO
40 TO
+200
40 TO
TO
"TO
7000
+2"'
,15
1.15
+200
+200
+2"'
+'"'
+""
,.,
U,
1.15
,.....
,
.
.......
.12
25
,45."
+2"'
-40
12.
.06
'12
.."
.06
LB2
+""
.. TO
+200
....
..
,....
'.0
112
647
SCR MANUAL
22.7 CIRCUIT ASSEMBLIES
SPECIFY GE SOLlD·STATE SUB-,ASSEMBLIES FOR A WIDE VARIID
OF CUSTOMER APPLICATIONS'
• m1abJe voJtap AC controls
• AC motor speed controls
• sialic switchilll
• _lta.e swltchln. AC power
controls
• temperature controls
• th,mron and hlp volta., rectifier
tube replacements
"5111"-STAIIIIMII TRIAC C1RCIIT SUIJ.ASSEM8IJES
•
_1I1IIIIp1f
l0 ...
15• ......
0IIISl1l12U.
2411lO11stlllSl.
... ..
_
_
....
.-_ a
_1111..........ues 11_ . ..,.. ..... \JpI11.2.
3~
If......-........ ...., ..
1fII,-....... ...-t _ _ _ " " -....
• AsclRlit _ _ - , . , . ,............. -
_
YllIo\J
....... .
lIlY ......
......... "'lind. SICII ..... _
............... _
- . . . -....... . . -.... ..,l1li11$.
lSI-I 1-1
-BUlB
ADE
1IIIe""': _ _
--.T
---==-
"
CCEgJ
,
~
~
=:.-.;:
=-=:=.
DICIfICm. iliff . . . . . . . _
648
1
t
-~-
....
II
II
2
I
t
5
'liIlIL,I./';Ir.'II.IIIllIo
DITJ1 EOlH FUB
... _----,
==-..,:==:.,-. ...
-- -- .,.1
•
111
1.
111
111
ll11J
-ll11J
ll11J
om
a-c _ _ _
GENERAL ELECTRIC THYRISTOR AND DIODE CONDENSED SPECIFICATIONS
"S200"-AC POWER CONTROLLERS
Futuns:
-
I.,....
• A........ 1IIfIp 0110 lid 15
(1115) 11120, 240 1111
m Vlns (IllS) I.. COIIInQlIII_ ...... fn115DD II 4150
•
• HIP Iopul nptCuce _
... 01 ....." IIIIfIIIIIIII ..,
IIIlIIfru5Kllll1111ollnL
• CtItnI polllllptlllbRIty bllfl, tIIItI ± 5.0% 01 ....., ...
CH_
-..
"ZIrI-VtIIIp SWItaIq" ............._ ...., RflI..... II1II
.. ,...1111. willi
_1cII
• AIIItIIHIIIICHlnI III l1li1111 IOd,1IIP roIilbllllJ.
• c.,.III. 01 optrlllot: willi I Vlrlt\y 01 nrllbl. ...IItH••
....." ..._ II thmbrllrs, plttll-nsl...". ~umldllJ lUllIIVIdnl.... ttc.
IIOIIEllCUTIIIE
S2IIOA
_IT
_TIC
c,
11[1.".....«
HrATE"
21
3
lOA
lOA
lOA
15&
.,
15&
15&
IDlTIIl
1m
241¥
271Y
12IY
241¥
2710
Contact.ICtG., ....., ...I.".
'-
lPEtIflCAnlN SlEET 110.111.40
"S300"-PHASE CONTROL POWER MODULES WITH FEEDBACK
I..,...." m_
11._
• Anil'" ill. rIIIn.. of .. 10111115
240 _ (HIS). fa, collll1llllnl
Ind. up II 3II1II willi.
(US) 11120
hlductlVl
or
•
AlllOlid-sllte.- fa, lui 1111 Ind _ip roIllbHIty.
CopIMI 01 opon1Inl willi I nrlt\y DI _ . _ .
..n.... )tIIemlsllr•• pIJoll-mistors, IHuIldlty ....Itin ...
_Fl. ttc.) II WIllI I. DC Ileh.mllt' ....1s.
Hltlti1qlatll1pe\lllcl .........I ..........tlRt:lI..,
pelll_51.,111111 ..... ,
hI_
• HIP pin 1rItP, cI...itry _
..... wIIh _I ......
.ptld CHlnI _Icati.n•.
"",I,
· _AI
• Vllllp-rtplated (;nner) •••trol .IrcIK pmid.. MIIIIJ in
VIItsp varialiDnL
•
•
• Ujl_. pl. allows m.dal...... In • notly .1 At mot.,
Wldl II", DI ClllllllIII
sIpII.
OII1p......d III ...~HIty dgrlnt; tn,eralln ft.....
ti.n•.
• 11_ Ind 1lIan: ...... g.. willi .., raslstlv. or IIgInI
_,flctor_INtIs.lli_OC ..m_ntsCl1llllllliJ
11.111 hi pmHtIlrDII.d _ I u d .lrnKL
"'q••..,
• Rldlt
ilia"'..... _IISSin network ••illim
.._ d ad IIdllltd Ifl.
_TI.
_II ._11
IA
~kbe~~
'''ClflUTIIIiI..u.'.I_
Il
III
I.
lOA
ISA
--
VGLlIIl
12"
l2IIV
l2IV
649
SCR MANUAL
"S400"-FANMOTOR SPEED CONTROLS
• lIII ...d.Is_.I4.......... 120_1J111S1.
• .,........ oIrnIt fir prj.." ..... fln""".od_
• AllooIld_ _ Ior""'I·.... 'IPIllIoIII1I\J.
• :._Itap..:~~_~IJ:~-=
•
Output ..riIIIoI "nlll~ II :I: 3 _
CopaIoIo of opllllloe In • wid. ~ of oppllcotll... wItIIIl
its I1IIq. uc, II; "III ai, ceodltio••rt. _
"'... III
... cellii' CIIIIIItionnn, exMuI fins. ull_.d l1lien,
po.p'••~ Clllal... _.IoIItI... Hd ......
"""1fIpI
.........
•
III 0111 hocl.d., a mola _
poIIntIo.,...... _
, . - - fir atlJntinlIIiII... output ..............
=':t.t,~H, ~ UL IPprmd priofotl ....ull
=:
••
_IIIC1AlUIIE
sc_DlAnes
SUfflll
IUIICTIOI
AI
AIS
AlSC
SJlEClflCAnONSIllER NIl. 111.10
TUBE REPLACEMENTS
StIId-slate IUb. IIplac_nts at mllabll la, hl&b...nop IICIIIIer tub... til,,","",••••tho, spacial pUIpIII tubes. Sile•••11
1Ip1s....." at fI' _tit purp..... II Is IbHlololJ
IfIot tillY be dlnct ....nical Ind .-itll IIplle_nb. Fo,
IIIIII·purp... tub,...ush II IIIJratro...
Is 11Il0l II
llsull IfIot till ..11d-state 1Ip1..._ , II In fact • 1111.......
IIIn' f" til. partlcollr _ _ oppIicItIol. CI... 1liiian Is
..hlIIIR" _
...ullllll' ........ oed or _ to ...11
_ill . .
._Ial
prapar 1Io·1an·
-
....
._or
"'. VI-
CIIIoHoCurnnt
T..,.
T•
826
200V
1.0A RM8 @1O"e
126°C
.D21
8'7
200V
l.oARMS@~oC
126°C
2060
I-E
.....
..:..
SPEClnCA'.......uTtcO. t8."
CUSTOMIZED ASSEMBLIES
,...11.
¥IIi'"
C.lDpletl ••mlo.._
ol",ull ......bll...... disOllll ....
or. mIIabI, In a
at _ _ ......_ ..
Includlll prlnlld .I",ull _ .. ptllld .odul... _lnI tub.
- . end mon, _lal pacbps to moot illllIl~..1
n.....
.uIIom"
650
SCH ad TRIAC .,lIor ope...DlllrDIL Sali~ l1l\I 1Ip1ac,_ for
1hJraIro. ItIIIn. Static switcbi... HIP fillip 1ICIiIIe, sfactIS.
Mold.d 1IIItfpl, d..... m_1or ."plll, I........uRL Molded
SCI .nd Inn.1sIIr .utes for ........ ad lIbor ..... T..·
pllllull ••nlllRo... AoIImItIc _"IIIIP _ I . lor . , .
_ills, 010. I.IPt _ d •. - .
GENERAL ELECTRIC THYRISTOR AND DIODE CONDENSED SPECIFICATIONS
22.8 RECTIFIER & SCR MODULES
• Most Standard Rectifier Diodes and SCR's, or CombInations of Both, Available with Hasl8lnks.
• A Wide Variety of Standard Circuit ConfiguraUons.
• Spacial Circuit Configurations Available Upon Re-
• Only Mounting Bolts and Electrical Connections tor
Power (Trigger Signal also, If SCR'. are used) are
Required.
• Proven Construcflon--GFs Long Years of Experience
W~t:: ~~dB:I~2~"I!h:::n.:s g.!.~~=~
quest
2 AMP BRIDGES-UP TO 1000v
• SIqIe , . . fin wm plrfol'lUllCt
• 4 IIIrmltlcllly ,,_ mllanclll protected A14 rHIHiIn
.1oi.. cllll.,1. ....
• -85·C II .... ,WC lperttiq ...,.rablre nIP
e2aps@2'·C;U.mps@WC
.7~trJn:
111110
IEl1I1
111112
111104
111111
H8111.
IIBIIO
5IIV
1DIV
2IIGV
.. _
....... _
. ....... _
I_
•=,"'r::.~:"CIIfIl\lmillU au
IPICIFICmOllItlftT ID. , .....
SILICON RECTIFIER MODULES-UP T03A, 12,.
A1 .....1AB1 I• • JDDY, 1.1A.
Itftg........ FuII . . . BI1...
Inl.lxJl5xe-PclnedBJoalL
eEl
CELL DATA
8-Ji!h
B
PACKAiE TYP..
E--C-EL-L....
·RAnNII
• A14'S
• 10DD VlCELL MAX
• lOA SINGLE CYCLE
BURGE
..
oB
M
N
50V
IOOV
200V
_
eoov
eoov
CIRCUIT
,......
-
*tnoludlllN171D-48
651
a'>
U'1
N
Medium & High Current Rectifier Modules
11
:::0
Cll1liA
B Ii
using A20 cells an type 11 Fins, capable of
conducting 6.3Aave.per cell at 1800 C
conduction angle in free air or 9.8A per cell
in 2000 linear feet per,minute forced air.
CELL
DATA
~
:z
~
121
13
NUMB.ER OF
CE'.LS
IN PARAlLEL
EACH LEG.
14
fll J:l" i"!'I~,r~f
~
•
•
:,~
~-:";"!':
~i;';~' ~r.!'
CEll MAX VOLT SINGlE~ FREE
200~(~E£
'20001FREE
,2000!fREE
2000 FREE
12000
NUMBER PER CEll SURGE-AMPS CONVECT LFPH ~VECT lFPM CONVECT LFPM CONVECT !LFPMICONVECT LFPM
A.
A20
600
150
6.3
O.
A25
600
240
9.' 15.5
Al8
400
150
A27
1200
2.0
A35
600
400
12.5 25.0
20.0 29.0
Al8
1200
500
12.5
25.0
20.0 29,0
A70
1000
1600
A90
1000
4500
9.8
6.3
MECHANICAL FEATURES
NO BRACKETS OR ~IOUNTING
FEET .
STANDARD
OTHERS AVAILABLE ON
SPECIAL ORDER.
J
9.8
12.5 18.4
60.4* 9.4*
~
NOMENCLATURE
A2011BCIAOl is a 20CV, single phase center tap,
POLARITY
A.
64,4199.41 78.21 96 •9
44.2 so.61
83.7
220
I
160
POSITI~.
NEGATIVE
I 250
* 3lt" X 3;," extrusion
CELL PEAK REVERSE VOLTAGE RATING
25V - U
SOV - F
lOOV - A
~
.'*
,
2aDV - B
300Y ~ C
400V· 0
H
S,ngte
Phase
"'attwa~e
1 Fin (Cell)
tnBet'cCI,cu,!
500V - E
BODV - N
600V· M
900V ~ T
700Y - 5 lOOOV - P
2000V - L
NOTE:
NUMBER OF CELLS IN
FOR PRY RATINGS NOT LISTED
USE MULTIPLE LETTERS
FOR ie, ~ PB'" 1200V
SERI ES
EACH
]
LEG
CIRCUIT DESIGNATOR
I
o
o-or- C
'"
:----0.
~
S'"9'.
Phase
Center
Tap
2F,ns(Cells)
In Bas'cC'reul!
E' "",.,
2Fms(Cells)
In Batie Circl,l'!
M
B
:?\
'\'<,
~
~~:~:
Bridge
4Fins(Celts)
tnBas'cClfcuil
~
'?;il~'
S'"9"
oc
-
Amp
Bfldge
~.
':'::;
4 Fins (Cells)
In Bss'cCi'cui!
Y
~.
o;...,J
Three
Phase
3Fins(Celts)
In Basic Ci,cu,l
-r--*
~----; ;~~::
~ f~ ~: F~utlwave
"'
_Bridge
6", _Jot; "6Fms(Cetls)
In Basic C"cU'!
X
S
$-0
•
S;,
Phase
<>-OH
~
6F,ns(Celts)
In BasleClfcuil
SPECtAL
Exampte
VoIlaee{"rade
A2011t'X239'
I·
Eng,neerln!t Number
Used Fo. All SpKlals
SCR Medium &High Current Modules
NOMENCLATURE
BlIAlI1lIA
e10 12BA1ADl is III 200V, full wave (back to back)
br1dgeconnected e10 cells on type 12 Fins.
cl~ble
of controll1ng4.? IIIIP average per SCR
in free air or 6.2 ·amp average per cell with 1000
linear feet per minute forced a1r.
11>
!2!
ir-
~
CELL
DATA
o
CELL MAX VOLT SINGLE .ov
NUMBER PER CEll SURGE·AMPS
Cl0
400
MECHANICAL FEATURES
NO BRACKETS OR MOUNTING
FEET.
D. STANDARD
OTHERS AVAILABLE ON
SPECIAL ORDER.
COMPATIBLE FREE
1000
RECTIFIER CONVECT LFPH
lN13414-46A
60
4.7
lN1341A-46A
A.
6.2
4.7
6.2
Cll
600
60
C35
BOO
ISO
1N2154-59
elSO
1300
1500
1N3292-96
44.0
CSO
SOO
1000
lN3289-92
52.0
62.9
64
78.1
060
400
1000
1N3289-91
Cl80
1300
3500
1N3135-42
e18S
SOO
3500
1N3735-39
4.7
6.2
10.1
16.3
3.98
6.3
6.0
20QY· B
300V-C
4DOY - 0
A.
B.
gJ
POSITIVE OR STANDARD
NEGATIVE
OTHER DEPENDING ON CIRCUIT
0'>
U1
""
S'
Songle Pllese Bridge
o
z
c
!2!
NUfoIBER OF CELLS IN
a'"
SERIES EACH LEG
FOR PRY RATJNGS NOT LISTED
USE MULTIPLE LEITERS:
FOR te. • PR· 1200V
S'
Smgle Fhase Bridge ISeR
.~ <5F~'" R'~'
'~~
'i
T
:s7
IMany AddillOl1al Varlll!lons Are Ava,lableUpol1 Requestl
I~
<5
c
(')
'"~
CIRCUIT DESIGNATOR
A
Z
C
c
1'T'1
i
NOTE:
~
»
POLARITY & MINOR ELECTRICALJ()DIFICATJONS
1 106
106
SOOV - E 8OOV· N 2000V· L
6OOV-H 900V-T
700Y - S lOaaV - p
-<
:::tl
:::tl
12.2
CELL PEAK REVERSE VOLTAGE RATING
25V - U
50Y-F
lOOV - A
-I
:::I:
F
Three Phase Fuliwave Bridge
(SCA's (;011)"1011 Cathode)
JWJ
H
'1
=f
Single Phase
Hallwave
X
SPECIAL (Followed By
Arbitrary fltJmber)
Eumple
vOllege,G.elle
CI0120X~39
E"9onee""gNymbe.
Uled for All Speciel~
::;;
~<5
z
'"
SCRMANUAl
22.9 SELENIUM COMPONENTS
pmI."""
..., 01 .... illllo..nt .....otaps II...,..,....... _ . . _ IiIIIHIJsIII
or silicon .... pal........, _
,..,.. yot, .. _lllppIiClliIIIs _ _...... _,......, VAC-IJ-sn*. solon;' nII. . . . . - _
"Uo ........ ,..11 HlflrmljOllllJ••f II1&II""- - - - . .f exceIlont ........ ofllllfltJ. l1li11 ~IP 1IIiINI\J.
HIGH-VOLTAGE MINIATURE CARTRIDGE
, ....,rete Iino of IDIfCt$!, mlnlllln soIon_ ..... eptImlzH IPplio.tion opportunities i. NiCad - . , - . . . . , ......h
••"m.... mo", ....d ••_
01_
.. II1II I.... dim......
nllillit _ . In•
....hlnl... prwpll1l0s. , ..Hable i•• vorl", of . - . _it .........
tio.slncludillg "'1I01lhllO. hllf....... conto"IIiI_rslllll ~ pIUs
spe...ls in c.11 sizes., to 15/32" ..II1II.
ffoxJ oncapsulated. tho.. m..iIIa...
RaiiRls and Specifications
Cu....1a.....
. ..... 2 IlIA to 150 IlIA
."pu1 Voila.. Ron...
.Up to 54Q _
(IIIIS)
Complete ratings and specifications available in Publication Numr 180.25.
.ENCAPSULATED MINIATURES
_rial••ip...1Iap lIit1it1ture . _.. (toWaoj _or dlodos ilea..,..
.... 111.. coli. will'" greatly I....... _ . ~ '" I at- ooit size.
cop and .p1IIJ-....odlJp. . . . . . . - .
AppIlcaIioas i..........1IIIrCiaI1IIII_.....-...paiaIIIt; .......
m.......pliooting .......... _ .._ ipitiol...,...., ...._ _
tors, ...........'.111 . . . .....
CSmol ............ "'.. U .. toUIllA, ... _
. . ._wiIII
PRY ratinp ., .'&11 .s 31.5110 volls. Complete !lUngs and specifications
available in Publicolion fiu"""" 1SO.50 and l8O.51.
_.nd
654
'"_
~_
GENERAL ElECTRIC THYRISTOR AND DIODE CONDENSED SPECIFICATIONS
THYRECTOR DIODES
rr...
EJoctric _ _ DiMes
iIII VoItIp Slppnssars) _
_ _ ... _rsapilst ............1tIp - . . . .
_
. . . . . . . .1Iios IS .........ppnssars fir plldlol.ialfHtYStll
MINIATURE THYRECTORS
_
iooillllr 1137!' or 15/32" ...... c.lls, Ira. 30 ta aao_ (IllS).
Complete ratings and specifications available in Publication Numbers 180.31
.... 180.36.
ARC SUPPRESSORS
A.-II1_ of llie CE VAC-IJ.SEL proCIS. praviol.. nillMo ..._ .
IsIIcs fir _
of tnoosieat ..11Ip map_. i••alnDid tirouits.
........ iol/37!' "'15/37!' ruund toll size•. Maximum DC supply .......
,... _
IIIooIIiaI all is 3D ..Its. Complete ratings and specificatians
avaYabI. in Publication Number 180.40.
EPOXY·ENCAPSULATED THYRECTORS
--- _d
r _ _ in In ......Itap
~lsIributiDn .,....... and o,.inate IIoIIt
_ ... _.SJS\IOI._protectl... d.mapislikolyla_,
ta oR - 1 I I I I s ; .....1IIIr SOIIieDnduetars, I..ps, cllck.mrs, l1li
Two m. ."'..... op..,...capSllated TbJreotar Diodos 1100 bo.. da.lped
IfICiIiaI1J fir _ _ '" bolllfllid _lucas, TV. and ...10 prDlllclln.
Dor ....... , . . - "mnunt ..1tIps 1Io1nl __
,..,._Iy.
..... IS . . . 2DDD . . . . . _ .
Complete ratings and
specifications _ . in Publicatian Numbers 180.33 and 180.34.
1" SQUARE THYRECTORS
TWI\J 1211 ..... "- 25 ta 5111_ (IllS). Tllnl&lHlllIIII ~ _ .
. . - - . ... _ _ _ ... _
Complete ratings and
specifications _
in Publicatiol> Numbers 180.30 and 180.35.
655
SCR MANUAL
··22.10 GENERAL ELECTRIC GE·MOVN METAL OXIDE VARISTORS
Description:
General Eiectric Metal Oxide ~ are voltage dependent, symmetrical resistors which .perform in a: manner similar to back-ta-back
zener diodes in circuit protective functions and offer advantages in
performance and economics. When exposed to high energy voltage
transients, the varistor impedance changes from a very high standby
value to a very low conducting value thus clamping the line voltage to a
safe ievel. The dangerous energy of tne incoming high voltage pulse is
absorbed by the GE-MOV varistor, thus protecting your voltage
sensitive circuit components.
I
GE-MOV VARISTORS-YOUR VOLTAGE TRANSIENT PROTECTION
I
a.ctrIcai Symbol
Rating Table (Maximum Values):
Storage Temperature, TSTG . . . . . . . . . . . . . . . . . . . . . . . . . . . . • . . . . . . . . -4O"C to +12S'C
Operating Surface Temperature, TS . . . . . . . . . . . . . . . . • . . . . . . . . . . . . . . . . . . . . . . IIS"C
Operating Ambient Temperature (without derating)
. . . . . . . . . . • . • . . . . . . • . . . . . . . . . 8S"C
Model
Number
VP130Al0
VP130A20
VP150Al0
VP150A20
VP250A20
VP250A40
VP420B4O
VP460B40
VP480B4O
VP480B80
VP510B4O
VP510B80
VP1000B80
VP1000BI60
RMS Input
Voltage
Volts
Recurrent
Peak Voltage
Volts
130
184
150
212
250
354
420
460
595
650
480
679
510
721
1000
1414
Energy
Rating
Joules
10
20
10
20
20
40
40
40
40
80
40
80
80
160
A_age Power
Dissipation Rating
Watts
0.5
0.85
0.5
0.85
0.6
0.9
0.9
0.9
0.7
1.0
0.7
1.0
0.9
1.3
Peak Current For
Pulses Less Than
7 Microseconds Wide
Amperes
1000
1250
1000
1250
1000
1250
1250
1250
1000
1250
1000
1250
1000
1250
Electrical Characteristics:
Model
Number
VP
130Al0
13OA20
150Al0
150A20
250A20
250A40
420B40
460B40
480B40
480B80
510B4O
510B80
loooB80
1oooB 160
Varistor Peak
Voltage
@lmAAC
(Peak)
Min.
Max.
Volts
Volts
184
249
212
287
354
479
595
650
805
880
679
914
721
968
1414
1900
Minimum Alpha*
1, = 1 mA
12=1 A
a (1-1000)
25
Capaeitance
(Typical)
Picofarads
1000
2000
1000
2000
700
1400
450
450
430
800
430
800
200
350
Maximum
Voltage
TempenitIITe
Coefficient
Body-to-Air
'IJ"C
"C/W
-0.05
Maximum
. Thermal
Resistenee
60
37
60
37
50
35
35
35
45
30
45
30
35
24
FOR COMPLETE SPECIFICATION SEE PUBLICATION 180.59
APPLICATION AND SPECIFICATION
23
L/TERATU~E;
SALES OFFICES
APPLICATION .AND SPECIFICATION LITERATURE;
SALES OFFICES
General Electric semiconductor Application Notes and specification sheets provide more detailed application and device specification
information than is possible in this Manual. Copies of the literature
listed in Section 23.1 and as indicated in Section 23.2 may be ordered
by Publication Number from:
Inquiry Clerk
General Electric Company
Semiconductor Products Department
Building 7 - Mail Drop 49
Electronics Park
Liverpool, New York 13088
USA
However, those Manuals listed as ERTM 3296, ETRM 3875A or
ETR 3960A must be ordered from:
General Electric Company
Department B
3800 North Milwaukee Avenue
Chicago, Illinois 60641
A check .or money order payable in .US dollars must accompany
each order.
Section 23.4 lists other General Electric Departments furnishing
related electrical and electronic components.
23.1 SEMICONDUCTOR DEVICE CATALOGS
Two versions of this catalog abound. One is the Short Form
Catalog (Publication No. 451.80). All of the devices appear in it with
their condensed. specifications. However, they have been arranged by
application so that the "right" device for the job may be easily selected
from among its brethren. Cross reference lists and selector guides
further facilitate this task.
A bound Semiconductor Data Handbook is available which has
,more comprehensive data on each device. The price of this 1000+
page Handbook is $3.95, which includes periodic updating.
Both these catalogs can be ordered from the address following in
Section 23.2.
23.2 APPLICATION NOTES
Publication Number 200.0, "Semiconductor Applications," contains abstracts of application notes, article reprints, technical papers
657
SCR MANUAL
and application manuals listed in the following sub-sections. Particular
publications which interest you may be ordered by publication number
from: Inquiry Clerk, General Electric Company, Semiconductor Products Dept., Bldg. #7 Mail Drop 49, Electronics Park, Liverpool, N. Y.
13088.
23.2.1 General Applications for Power Semiconduct1Jrs
90.16
90.21
90.57
90.58
90.68
90.83
90.44
200.1
200.5
200.9
200.10
200.15
200.19
200.28
200.30
200.32
200.34
200.35
200.36
200.38
200.39
200.42
200.50
200.55
201.23
660.13
660.14
660.15
660.16
660.21
671.1
671.12
658
Silicon Controlled Switches
How to Suppress Rate Effect in PNPN Devices
Using the Silicon Bilateral/Unilateral Switch
Reversible Ring Counter Utilizing the Silicon Controlled
Switch
The Silicon Unilateral Switch Provides Stable, Economical
Frequency Division
A highly Reliable, Fail Safe, Precision Undervoltage
Protection Circuit
The Complementary SCR
Characteristics of Common Rectifier Circuits
General Electric Selenium Thyrector Diodes
Power Semiconductor Ratings Under Transient and
Intermittent Loads
Overcurrent Protection of Semiconductor Rectifiers
Turn Off Time Characterization and Measurement of Silicon
Controlled Rectifiers
Using Low Current SCR's
The Rating of SCR's When Switching Into High Currents
Capacitor Input Filter Design With Silicon Rectifier Diodes
A Variety of Mounting Techniques for Pres-Fit SCR's
and Rectifiers
The Light Activated SCR
Using the Triac for Control of AC Power
The Solid State Thyratron
Application of Fast Recovery Rectifiers
The Series Connection of Rectifier Diodes
Commutation Behaviour of Diffused High Current
Rectifier Diodes
Mounting Press Pak Semiconductors
Thermal Mounting Considerations for Plastic Power
Semiconductor Packages
SCR - Ignitron Comparison
The Rating and Application of SCR's Designed for Power
Switching at High Frequencies
Basic Magnetic Functions in Converters and Inverters
Including New Soft Commutation
SCR Inverter Commutated by an Auxiliary Impulse
An SCR Inverter with good Regulation and Sine-Wave Output
Take the Guesswork Out of Fuse Selection
Economy Power Semiconductor Applications
Optimum Solid State Control Parameters for Improved
Performance of In-Space Electric Heating Systems
APPLICATION AND SPECIFICATION LITERATURE; SALES OFFICES
ETRM-3875 Silicon Controlled Rectifier Manual, 5th Ed., $3.00
ETRM-3960 G-E Electronic Experimenter Circuit Manual ($2.00)
23.2.2 Silicon Controlled Rectifier and Other Thyristor Circuits
200.18
200.21
200.31
200.33
200.43
200.44
200.47
200.48
200.49
200.53
200.54
200.58
201.1
201.6
201.9
201.10
201.11
201.12
201.13
201.14
201.15
201.16
201.17
201.18
201.24
Fluore~cent
Lamp Dimming With SCR's and Associated
Semcionductors
Three Phase SCR Firing Circuits for DC Power Supplies
Phase Control of SCR's With Transformer and Other
Inductive AC Loads
Regulated Battery Chargers Using the Silicon Controlled
Rectifier
Solid State Control for DC Motors Provides Variable
Speed With Synchronous - Motor Performance
Speed Control for Shunt-Wound Motors
Speed Control for Universal Motors
Flashers, Ring Counters and Chasers
A Low-Cost, Ultrasonic-Frequency Inverter
Using A Single SCR
Solid-State Incandescent Lighting Control
Design of Triggering Circuits for Power SCR's
Solid State Electric Heating Controls
A Plug-In Speed Control for Standard Portable
Tools and Appliances
Touch Switch or Proximity Detector
Precision Temperature Controller
Auto, Boat, or Barricade Flasher
Time-Delay Relay
500 Watt AC Line Voltage and Power Regulator
Universal Motor Control With Built-in Self-Timer
Two Automatic Liquid Level Controls
Solid-State Control for Electric Blankets
Fan Motor Speed Control- "Hi-Intensity" Lamp Dimmer
Sequential Turn Signal System for Automobiles
High Voltage Power Supply for Low Current Applications
Thyristor Selection for Incandescent Lamp Loads
\
23.2.3 Unijunction Applications
90.10 The Unijunction Transistor Characteristics and Applications
90.12 Unijunction Temperature Compensation
90.19 Unijunction Frequency Divider
90.70 The D13T - A Programmable Unijunction Transistor
90.72 Complementary Unijunction Transistors
23.2.4 Test Circuits
201.3
Portable SCR and Silicon Rectifier Tester
659
SCR MANUAL
23.3 SPECIFICATION SHEETS
The device specification sheets referred to on the condensed
specifications of Chapter 22, and others that may be mentioned in this
Manual, may be ordered by Publication Number from:
Inquiry Clerk
General Electric Company
Semiconductor Products Department
Building 7 - Mail Drop 49
Electronics Park
Liverpool, New York 13088
USA
23.4 RELATED GENERAL ELECTRIC DEPARTMENTS
The Semiconductor Products Department would like to remind
its readers of the great variety of parts and services that our sister
departments in General Electric can furnish to meet your SCR circuit
needs: Among them are:
- SCR Capacitors for phase control and inverter commutation
duty (GE Industrial and Power Capacitor Products Dept.,
Hudson Falls, N. Y.
- Transformer and chokes for inverters and power supply needs
(GE Specialty Transformer Dept., Ford Wayne, Ind.)
- A complete family of GE cadmium sulfide cells and magnetic
reed switches (GE Tube Products Dept., Owensboro, Ky.)
- A wide variety of electronic capacitors and rechargeable nicklecadmium batteries (Electronic Capacitor and Battery Products
Department, Irmo, S. C.)
- A complete family of high-reliability lamps and light emitting
diodes for activation of light-activated SCR's (GE Miniature
Lamp Dept., Cleveland, Ohio)
- Silicone potting and joint compounds (GE Silicone Products
Dept., Waterford, N. Y.)
- And, of course, the industry's most complete line of motors and
other related electrical and. electronic equipment
23.5 GENERAL ELECTRIC SALES OFFICES
All products of the Industrial ,and Power Capacitor Products,
Electronic Capacitor and Battery Products, Semiconductor Products,
Specialty Transformer and Tube Products Departments' as well as the
light emitting diodes of the Miniature Lamp Department are sold
through General Electric's Electronics Components Sales' Department
(ECSD) and through Component Sales Department (CSD) to original
equipment manufacturers (OEM's) and distributors of electrical/electronic equipment and components. OEM Sales Offices of ECSD and
CSD are listed below:
660
APPLICATION AND SPECIFICATION LITERATURE; SALES OFFICES
State
Electronic Components
Sales Department
Binningham 35202
2151 Highland Ave.
P.O. Box 2602
205 322-7683
Alabama:
Arizona:
Phoenix 85012
United Bank Bldg.
3550 N. Central Ave.
602 264-1751
Phoenix 85012
United Bank Bldg.
3550 N. Central Ave.
602 264-1751
North Little Rock 72119
2nd and Main Sts.
P.O. Box 5641
501 376-3458
Arkansas:
California:
Component
Sales Department
Burlingame 94010
Los Angeles 90064
11840 W. Olympic Blvd. 25 Edwards Court
213 272-8566
415 692-0700
Portola Valley 94025
3210 Alpine Rd.
415 854-4010
Los Angeles 90015
1543 W. Olympic Blvd.
213 381-1247
Colorado:
Denver 80206
201 University Blvd.
P.O. Box 2331, 80201
303 388-5771
Denver 80206
201 University Blvd.
P.O. Box 2331, 80201
303 388-4545
Connecticut:
Bridgeport 06602
1285 Boston Ave.
203 334-1012
Meriden 06450
1 Prestige Dr.
P.O. Box 910
203 238-0791
District of
Columbia:
Washington 20005
777-14th St., NW
202 393-3600
Florida:
North Palm Beach 33403 Tampa 33609
321 North Lake Blvd.
2106 S. Lois Ave.
305 844-5202
P.O. Box 10577
813 877-8311
Tampa 33609
2104 S. Lois Ave.
P.O. Box 10577
813 877-8311
Winter Park 32789
John Hancock Bldg.
370 Wymore Rd.
305 647-2030
661
SCR MANUAl
State
ECSD
CSD
Georgia:
Atlanta 30329
1699 Tully Circle, N.R
404 633-4522
Illinois:
Oak Brook 60521
Chicago 60641
3800 N. Milwaukee Ave. Oakbrook North
1200 Harger Rd.
312 777-1600
317 654-2960
Indiana:
Ft. Wayne 46806
6001 S. Anthony Blvd.
219 447-1511
Indianapolis 46208
. 3750 N. Meridian St.
317 923-7221
Atlanta 30309
1860 Peachtree Rd., N.W.
P.O. Box 4659,30302
404 351-4400
Evansville 47714
2709 Washington Ave.
P.O. Box 3357, 47701
812 477-8821
Ft. Wayne 46804
1635 Broadway
Bldg. 18-5
219 743-7431
Indianapolis 46240
55 Winterton
1010 E. 86th St.
P.O. Box 40216
317 846-6564
South Bend 46601
430 N. Michigan St.
219 234-4196
Iowa:
Cedar Rapids 52401
210 Second St., SE
303 Dows Bldg.
319 364-9149
Bettendorf 52722
2435 Kimberly Rd.
314 359-0351
Des Moines 50322
7200 Hickman Rd.
P.O. Box 3809
Urbandale Branch
515 278-0451
Kansas:
Overland Park 66204
7219 MetcaH Ave.
Mailing Address:
Shawnee Mission 66201
P.O. Box 408
913 262-0442
Louisiana:
662
New Orleans 70112
National Bank Bldg.
613 Hibernia
504 525-4324
APPLICATION AND SPECIFICATION LITERATURE; SALES OFFICES
State
ECSD
CSD
Massachusetts:
Wellesley 02181
1 Washington St.
617 237-2050
Wellesley 02181
1 Washington St.
617 237-2050
Michigan:
Southfield 48075
22255 Greenfield Road
313 355-4400
Grand Rapids 49508
2821 Madison Ave., S.E.
P.O. Box 710
616 452-2121
Southfield 48075
Box 1316
Northland Center Station
313 872-2600
Minnesota:
Minneapolis 55442
4900 Viking Dr.
612 927-5458
Minneapolis 55435
4018 West 65th St.
612 927-8814
Missouri:
Kansas City 64199
911 Main St.
P.O. Box 13566
816 221-4033
St. Louis 63101
1015 Locust Street
314 436-4343
St. Louis 63132
1530 Fairview
314 429-6941
New Jersey:
Clifton 07014
200 Main Ave.
201 472-8100
East Orange 07017
56 Melmore Gardens
201 675-9426
New York:
Albany 12205
11 Computor Dr., West
518 458-7755
Mattydale 13211
5858 E. Malloy Rd.
315 456-7432
East Syracuse 13057
7 Adler Dr.
315 456-1046
Great Neck 11021
425 Northern Blvd.
516 466-8800
Rochester 14618
3380 Monroe Ave.
P.O. Drawer C
12 Corners Branch
716 586-6474
Rochester 14624
35 Deep Rock Rd.
716 436-3480
663
SCRMANUAL
state
EGSD
North Carolina: Charlotte 28211
2915 Providence Rd.
704 364-6313
GSD
Charlotte 28207
141 Providence Rd.
P.O. Box 1969,28207
704 375-5571
Greensboro 27408
1828 Banking St.
P.O. Box 9476
919 273-6982
Ohio:
Cincinnati 45206
2621 Victory Pkwy
513 281-2547
Cincinnati 45206
2621 Victory Pkwy
513 861-3400
Cleveland 44117
25000 Euclid Ave.
216 266-2900
Cleveland 44116
20950 Center Ridge Rd.
216 333-0552
Dayton 454439
3430 S. Dixie Hway
P.O. Box 2143
Kettering Branch 45429
513 '298-0311
Dayton 45439
3430 S.Dixie Hway
P.O. Box 2143
Kettering Branch 45429
513 298-0311
MansBeld 44902
166 Park Ave., West
419 524-2622
Toledo 43606
3450 West Central Ave.
419 531-8943
Oklahoma:
Oregon:
664
3315 E. 47th Place
Oklahoma City 73Il2
3022 Northwest EXpway Tulsa 74135
Suite 100
405 943-9015
918 743-8451
Portland 97210
2929N.W.29thAve.
P.O. Box 909, 97207
503 228-0281
APPLICATION AND SPECIFICATION LITERATURE; SALES OFFICES
State
Pennsylvania:
ECSD
CSD
Erie 16505
2318 West 8th St.
814 455-8377
Fort Washington 19034
1260 Virginia Drive
215 643-1633
Philadelphia 19102
3 Penn Center Plaza
215 568-1800
Pittsburgh 15234
Lebanon Shops
300 Mt. Lebanon Blvd.
412 531-6655
Pittsburgh 15220
875 Greentree Rd.
3 Parkway Center
412 921-4134
York 17403
1617 E. Market St.
717 848-2828
Tennessee:
Chattanooga 37411
5800 Bldg., Eastgate Ctr.
615 894-2550
Nashville 37204
2930 Sidco Dr.
615 254-1187
Texas:
Dallas 75205
Dallas 75247
4447 N. Central Expway 8101 Stemmons Freeway
214 521-1931
P.O. Box 5821, 75222
214 631-3110
Houston 77006
Houston 77027
3110 Southwest Freeway 4219 Richmond Ave.
713 524-3061
P.O. Box 22045
713 623-6440
Virginia:
Charlottesville 22903
2007 Earhart St.
P.O. Box 319
703 296-8118
Portsmouth 23707
810 Loudoun Ave.
P.O. Box 7135
703 484-3521 ext 628
Washington:
Seattle 98188
225 Tukwila Pkwy
206 244-7750
Wisconsin:
Milwaukee 53202
615 E. Michigan St.
414 271-5000
Milwaukee 53226
Mayfair Plaza
2421 N. Mayfair Rd.
414 778-0259
665
SCR MANUAL
23.6 INTERNATIONAL GE SALES OFFICES
In Canada, address inquiries to:
Canadian General Electhc Co.
189 Dufferin St.
Toronto, Ontario, Canada
416534-6311
ENGLAND
International General Electric
Company of NewYork,Ltd.
Lincoln House
296-302 High Holborn
London W. C. I
Telephone 01-242-6868
SPAIN
International General Electric
Company of Spain, S.A.
Apartado 700
Avenida Jose Antonio
Madrid
Telephone 247.16.05
JAPAN
General Electric Japan, Limited
11-41, l-chome
Akasaka, Minato-ku
Tokyo
Telephone 582-0371
MEXICO
General Electric de Mexico, S.A.
Apartado 53-983
Marina Nacional No. 365
Mexico 17 D.F.
Telephone 545-63-60
ITALY
Compagnia Generale di Elettricita
S.p.A.
Via F. Casita 44
Milan
Telephone 63-93-64
SWEDEN
International General Electric AB
Gardsfogdevagen 14, II
161 70 Bromma
Telephone 28-29-45
GERMANY
General Electric-Germany
Postfach 3011
Eschersheimer Landstrasse 60-62
6 Frankfurt/Main I
Telephone 6II-1564-35
General Electric-Germany
Hermann Lingg Strasse 12
Munich 15
Telephone 537970
FRANCE
International General Electric
France, S.A.
42 Avenue Montaigne
Paris-8e •
Telephone 225-52-32
AUSTRALIA
Australian General Electric Pty.
Ltd.
103 York Street
Sydney, N.S.W., 2000
Telephone 29-87II; 29-7553
23.7 GENERAL ELECTRIC SEMICONDUCTOR DISTRIBUTORS
The distributor houses listed below have in stock the broad line
of GE semiconductors. They also carry many other GE electronic components so that you may secure. a wide variety of parts on a single
order from a single source. Also don't forget to consult the "Yellow
Pages" of your telephone directory for possible address changes or new
additions.
666
APPLICATION AND SPECIFICATION LITERATURE; SALES OFFICES
ALABAMA
CRAMERIEW HUNTSVILLE
Huntsville, (205) 539-5722
FORBES DISTRIBUTING CO.
Birmingham, (205) 251-4104
ARIZONA
HAMILTON/AVNET
ELECTRONICS
Phoenix, (602) 269-1391
KIERULFF ELECTRONICS, INC.
Phoenix, (602) 273-7331
ARKANSAS
CARLTON-BATES CO.
Little Rock, (501) 375-5375
CALIFORNIA
BRILL ELECTRONICS
Oakland, (415) 834-5888
CRAMER/LOS ANGELES
Glendale, (213) 243-6224
CRAMER/SAN FRANCISCO
Redwood City, (415) 365-4000
ELECTRONIC SUPPLY
Riverside, (213) 683-8110
ELMAR ELECTRONICS
Mountain View,(415) 961-3611
HAMILTON/AVNET
ELECTRONICS
Culver City, (213) 870-7171
Mountain View, (415) 961-7000
San Diego, (714) 279-2421
KIERULFF ELECTRONICS, INC.
Los Angeles, (213) 685-5511
Palo Alto, (415) 968-6292
San Diego, (714) 278-2112
G. S. MARSHALL CO.
El Monte, (203) 686-0141
San Diego, (714) 278-6350
WESTERN RADIO & TV SUPPLY
San Diego, (714) 239-0361
COLORADO
DENVER WALKER
ELECTRONICS
Denver, (303) 935-2401
ELECTRONIC PARTS CO.
Denver, (303) 266-3755
HAMILTON/AVNET
ELECTRONICS
Denver, (303) 433-8551
CONNECTICUT
BOND RADIO ELECTRONICS
INC.
'
Waterbury, (203) 753-1184
CRAMER/CONNECTICUT
North Haven, (203) 239-5641
FLORIDA
CRAMER/EW HOLLYWOOD
Hollywood, (305) 923-8181
CRAMER/EW ORLANDO
Orlando, (305) 841-1550
HAMILTON/AVNET
Hollywood, (305) 925-5401
HAMMOND ELECTRONICS
Orlando, (305) 241-6601
SCHWEBER ELECTRONICS
Hollywood, (305) 927-0511
GEORGIA
CRAMER/EW ATLANTA
Atlanta, (404) 451-5421
JACKSON ELECTRONICS CO.
Atlanta, (404) 355-2223
ILLINOIS
ELECTRONIC DISTRIBUTORS
INC.
'
Chicago, (312) 283-4800
HAMILTON/AVNET
ELECTRONICS
Schiller Park, (312) 678-6310
NEWARK ELECTRONICS CORP.
Chicago, (312) 638-4411
SEMICONDUCTOR SPECIALISTS
INC.
'
Elmhurst Industrial.Park,
(312) 279-1000
INDIANA
FT. WAYNE ELECTRONICS
Ft. Wayne, (219) 742-4346
GRAHAM ELECTRONICS, INC.
Indianapolis, (317) 634-8486
HUTCH AND SON
Evansville, (812) 425-7155
SEMICONDUCTOR SPECIALISTS
INC.
'
Indianapolis, (317) 243-8271
667
SCR MANUAL
IOWA
DEECO, INC.
Cedar Rapids, (319) 365-7551
KANSAS
HAMILTON/AVNET
ELECTRONICS
Prairie Viilage, (913) 362-3250
INTERSTATE ELECTRONICS
SUPPLY CORP.
Wichita, (316) 264-6318
KENTUCKY
P. I. BURKS CO.
Louisville, (502) 583-2871
LOUISIANA
EPCOR
New Orleans, (504) 486-7441
RALPH'S OF LAFAYETTE
Lafayette, (318) 234-4507
STERLING ELECTRONICS, INC.
New Orleans, (504) 522-8726
MAINE
HOLMES DISTRIBUTORS, INC.
Portland; (207) 774-5901
MARYLAND
MICHIGAN
NEWARK,-INDUSTRIAL
ELECTRONICS CORP.
Grand Rapids, (616) 452-1411
RS ELECTRONICS
Detroit, (313) 491-1012
SEMICONDUCTOR SPECIALISTS
INC.
'
Detroit, (313) 255-0300
MINNESOTA
GOPHER ELECTRONICS CO.
St. Paul, (612) 645-0241
LEW BONN COMPANY
Edina, (612) 941-2770
SEMICONDUCTOR SPECIALISTS
INC.
'
Minneapolis, (612) 866-3434
MISSISSIPPI
ELLINGTON ELECTRONICS
SUPPLY, INC.
.
Jackson, (601) 355~0561
MISSOURI
CRAMER/EW BALTIMORE
HAMILTON/AVNET
Baltimore, (301) 354-0100
ELECTRONICS
CRAMER/EW WASHINGTON
Hazelwood, (314) 731-1144
Gaithersburg, (301) 948-0110
L COMP-KANSAS CITY
HAMILTON/AVNET
North Kansas City, (816) 221-2400
ELECTRONICS,
NORMAN ELECTRONIC SUPPLY
Hanover, (301) 796-5000
Joplin, (417) 624-0368
KANN-ELLERT ELECTRONICS, INC. OLIVE INDUSTRIAL
ELECTRONICS
Baltimore" (301) 889-4242
PIONEER/WASHINGTON,
University City, (314, B63-4051
Rockville, (301) 424-3300,
RADIO LA'S., INC.'
SCHWEBER ELECTRONICS
Kansas City, (816)561-9935
Rockville, (301) 4,27-4977
NEBRASKA
MASSACHUSETTS
RADIO EQUIPMENT CO.
CRAMER ELECTRONICS, INC.
Omaha, (402) 341-7700'
Newton, (617) 969-7700
SCOTT ELECTRONIC
HAMILTON/AVNET
SUPPLY CORP.
Burlington, (617) 272-3060
Lincoln, (402) 434-8308 '
T. F. CUSHING, INC ..
Springfield, (413) 788-7341
GERBER ELECTRONICS
Dedham, (617) 329-2400
SCHWEBER ELECTRONICS
Waltham, (617) 891-8484
668
APPLICATION AND SPECIFICATION LITERATURE; SALES OFFICES
NEW JERSEY
CRAMER/NEW JERSEY
Union, (201) 687-7870
CRAMER/PENNSYLVANIA
Pennsauken, (609) 662-5061
GENERAL RADIO SUPPLY
CO., INC.
Camden, (609) 964-8560
HAMILTON / AVNET
ELECTRONICS
Cedar Grove, (201) 239-0800
Cherry Hill, (609) 662-9337
NEW MEXICO
CRAMER/NEW MEXICO
Albuquerque, (505) 265-5767
ELECTRONICS PARTS CO.
Albuquerque, (505) 265-8401
KIERULFF ELECTRONICS, INC.
Albuquerque, (505) 268-3901
NEW YORK
ARROW ELECTRONICS, INC.
Farmingdale, (516) 694-6800
CRAMER/BINGHAMTON
Binghamton, (607) 754-6661
CRAMER/LONG ISLAND
Hauppauge, (516) 231-5600
CRAMER/ROCHESTER
Rochester, (716) 275-0300
CRAMER/SYACUSE
Syracuse, (315) 437-6671
HAMILTON ELECTRO SALES
Syracuse, (315) 437-2641
Westbury, (516) 333-5800
ROCHESTER RADIO
SUPPLY CO.
Rochester, (716) 454-7800
ROME ELECTRONICS, INC.
Rome, (315) 337-5400
SCHWEBER ELECTRONICS
Westbury, L.I., (516) 334-7474
STANDARD ELECTRONICS,
INC.
Buffalo, (716) 883-5000
Endicott, (607) 754-3102
VALLEY INDUSTRIAL
ELECTRONICS, INC.
Yorkville, (315) 736-3393
NORTH CAROLINA
CRAMER/EW WINSTON-SALEM
Winston-Salem, (919) 725-8711
DIXIE RADIO SUPPLY CO.
Charlotte, (704) 377-5413
SOUTHEASTERN RADIO
SUPPLY CO., INC.
Raleigh, (919) 828-2311
OHIO
ELECTRONICS MARKETING
CORP.
Columbus, (614) 299-4161
HUGHES-PETERS, INC.
Cincinnati, (513) 351-2000
Columbus, (614) 294-5351
PIONEER-CLEVELAND
Cleveland, (216) 587-3600
PIONEER/DAYTON
Dayton, (513) 236-9900
REM ELECTRONICS SUPPLY
Warren, (216) 399-2777
SUN RADIO CO., INC.
Akron, (216) 434-2171
WARREN RADIO CO.
Toledo, (419) 448-3364
OKLAHOMA
OIL CAPITOL ELECTRONICS
CORP.
Tulsa, (918) 836-2541
TRICE WHOLESALE
ELECTRONICS
Oklahoma City, (405) 524-4415
PENNSYLVANIA
ALMO ELECTRONICS CORP.
Philadelphia, (215) 676-6000
RESCO OF LEHIGH VALLEY
Allentown, (215) 435-6743
ROSEN ELECTRONICS CO.
York, (717) 843-3875
R. P. C. ELECTRONICS
Pittsburgh, (412) 782-3770
SEMICONDUCTOR
SPECIALISTS, INC.
Pittsburgh, (412) 781-8120
RHODE ISLAND
W. H. EDWARDS CO.
Warwick, (401) 781-8000
669
SCR MANUAL
SOUTH CAROLINA
DIXIE RADIO SUPPLY CO.,
INC.
Columbia, (803) 253-5333
Greenville, (803) 239-1328
UTAH
KIMBALL ELECTRONICS
Salt Lake City, (801) 328-2075
NEWARK ELECTRONICS
Salt Lake City, (801) 486-1048
TENNESSEE
BLUFF CITY
DISTRIBUTING CO.
Memphis, (901) 276-4501
ELECTRA DISTRIBUTING CO.
Nashville, (615) 255-8444
HARPE ELECTRONIC
DISTRIBUTORS, INC ..
Chattanooga, (615) 267-2381
RADIO ELECTRIC
SUPPLY CO.
Kingsport, (615) 247-8111
VIRGINIA
MERIDIAN ELECTRONICS, INC.
Rich.-nond, (703) 353-6648
PEOPLES RADIO & TV
SUPPLY CO.
Roanoke, (703) 342-8933
VIRGINIA RADIO SUPPLY CO.
Charlottesville, (703) 296-4184
TEXAS
HAMILTONIAVNET
ELECTRONICS
Dallas, {214) 638-2850
Houston, (713) 526-4661
McNICOL, INC.
EI Paso, (915) 566-2936
MIDLAND SPECIALTY CO.
EI Paso, (915) 533-9555
NORVELL ELECTRONICS
Dallas, (214) 357-6451
STERLING ELECTRONICS
Dallas, (214) 357-9131
Houston, (713) 623-6600
WHOLESALE ELECTRONIC
SUPPLY
Dallas, (214) 824-3001
670
WASHINGTON
ALMAC/STROUM
Seattle, (206) 763-2300
C&G ELECTRONICS CO.
Tacoma, (206) 272-3185
HAMILTON/AVNET
ELECTRONICS
Seattle, (206) 624-5930
WEST VIRGINIA
CHARLESTON ELECTRICAL
Charleston, (304) 348-5211
WISCONSIN
ELECTRONIC
EXPEDITORS, INC.
Milwaukee, (414) 374-6666
MARSH RADIO SUPPLY CO.
West Allis, (414) 545-6500
APPLICATION INDEX
APPLICATION INDEX·
The circuits referred to in the following figure numbers are
intended as a starting point for the equipment designer in achieving
the detailed requirements of his application. Since these circuits are
not necessarily "ultimate" for every application, it is hoped the imaginative designer will use them simply as a jumping-ofI point for his
own development. Likewise, many of these circuits can be used for
other functions besides those mentioned in the text. As a guide to
some of the various thyristor circuits for accomplishing specific tasks,
here is a tabulation of figures in this manual classified by possible
application (please note that these are Figure numbers and not section
or paragraph numbers):
Applications
For Basic Circuit Possibilities
See Filure Number
AC Static Switches ........ 4.20, 4.27, 6.3, 6.10, 6.11, 6.12, 6.25,
6.26,6.27,6.28,7.7,7.11,7.12,7.13,8.1,
8.2, 8.3, 8.4, 8.5, 8.6, 8.7, 10.21, 11.3,
11.4, 11.5, 11.6, 11.7, 11.8, 11.9, 11.17,
14.24, 14.25, 14.28, 14.29, 14.30, 14.31
Appliance Controls ........ 4.20, 4.21, 4.22, 4.43, 4.44, 7.7, 7.11,
7.12, 7.13, 7.14, 8.1, 8.4, 8.5, 8.6, 8.23,
8.24, 8.25, 8.26, 8.28, 8.29, 8.30, 8.37,
11.6, 11.11, 12.6, 12.11, 12.17, 12.24,
14.35, 14.36, 14.37, 14.38
Battery Chargers ......... 4.25, 6.25, 6.26, 6.27, 6.28, 8.9
Circuit Breakers .......... 5.8,5.9,6.25,6.26,6.27,6.28,8.18
Current Regulators " ..... 6.25, 6.26, 6.27, 6.28, 9.40, 9.41
DC Static Switches ........ 5.8, 5.9, 5.11, 5.13, 6.3, 6.10, 6.11, 6.12,
6.25,6.26,6.27,6.28,8.10,8.11
DC to AC Inverters ........ 4.12,4.49,4.50,5.8, 5.9, 5.10,5.11,5.12,
5.13, 5.14, 5.15, 5.16, 5.17, 5.18, 5.20,
6.25, 6.26, 6.27, 6.28, 13.2, 13.6, 13.7,
13.8, 13.10, 13.11, 13.15, 13.16, 13.17,
13.23, 13.25, 13.28, 13.32
DC to DC Converters ...... 4.12, 4.49, 4.50, 5.8, 5.9, 5.10, 5.11, 5.12,
5.13, 5.14, 5.20, 6.25, 6.26, 6.27, 6.28,
13.2, 13.11, 13.15, 13.16, 13.17, 13.23,
13.25, 13.28, 13.32
DC Power Supplies ........ 4.25, 6.3, 6.10, 6.11, 6.12, 6.25, 6.26,
6.27,6.28
Driver Circuits ........... 8.39, 8.40
Electric Vehicle Drives ..... 4.12, 5.8, 5.9, 5.10, 5.11, 5.13, 5.14, 5.20,
13.2, 13.10, 13.11, 13.15, 13.16, 13.17,
13.23, 13.25, 13.28
Electronic Crowbars ....... 8.20
Exciters for Motors &:
Generators ............. 4.25, 6.3, 6.10, 6.11, 6.12, 6.25, 6.26,
6.27,6.28, 14.33
671
SCR MANUAL
Applications
F.r Basic Circuit Possibilities
See Figure Number
FlasherS ................. 4.49, 7.14, 8.12, 8.13, 8.14, 8.15, 8.16,
14.39
Firing Circuits ........... (See Triggering Circuits)
Flip Flops ............... 4.49, 4.50, 4.52
Frequency Changers
.. 4.12,4.52, 5.8, 5.9, 5.10, 5.11, 5.12, 5.13,
5.14, 11.19, 12.16, 12.17, 13.2, 13.10,
13.11, 13.15, 13.16, 13.17, 13.23, 13.25,
13.28, 13.32, 13.44, 13.46
Ignition Firing ........... 6.3, 6.10, 6.11, 6.12
Induction Heaters ........ 4.12, 5.9, 5.12, 6.25, 6.26, 6.27, 6.28,
13.2, 13.10, 13.11, 13.15, 13.16, 13.17,
13.25, 13.28, 13.32
Lamp Dimmers .......... 4.21, 4.22, 4.23, 4.26, 4.43, 7.7, 7.15,
12.11
Latching Relays .......... 14.14
Lighting Circuits ......... 8.36,9.44, 13.2, 14.42
Logic Circuits ........... 14.34
Motor Contro1s &
Drives-DC ........... 4.12, 4.25, 6.3, 6.10, 6.25, 6.26, 6.27,
6.28, 10.1, 10.3, 10.4, 10.5, 10.6, 10.7,
10.8, 10.9, 10.10, 10.11, 10.12, 13.15,
13.16, 13.17, 13.23
Motor Contro1s &
Drives - AC ........... 4.12, 5.10, 5.11, 5.12, 5.20, 6.25, 6.26,
6.27,6.28,7.7,10.1,10.14,10.15, 10.11,
10.18, 10.19, 10.20, 10.21, 12.11, 12.23,
12.24, 12.25, 12.26, 13.10, 13.11, 13.15,
13.16, 13.17, 13.23, 13.25, 13.28, 14.40
Oscillators ............... 4.32, 4.47, 7.14, 13.2, 13.32
Phase Controls ........... 4.21, 4.22, 4.23, 4.25, 4.26, 4.27, 4.43,
4.44, 6.3, 6.10, 6.11,6.12, 7.7, 7.15, 9.1,
9.14, 9.15, 9.17, 9.18, 9.20, 9.24, 9.25,
9.29, 9.30, 9.31, 9.32, 9.33, 9.34, 9.35,
9.39, 9.45, 9.46, 9.51, 9.52, 9.58, 12.6,
12.7, 12.10, 12.11, 12.24, 14.31, 14.32,
14.33, 15.18
Photoelectric Circuits ...... 6.11, 6.12, 11.17, 14.23, 14.24, 14.25,
14.26, 14.27, 14.28, 14.29, 14.30, 14.31,
14.32, 14.33, 14.34, 14.35, 14.36, 14.37,
14.38, 14.39, 14.40, 14.42, 14.43, 14.44,
14.45, 14.46
Power Supplies .......... 4.25, 6.3, 6.10, 6.11, 6.12, 6.25, 6.26,
6.27, 6.28, 7.7, 9.45, 9.46, 9.51, 9.52,
9.53,9.60,9.61, 11.9, 11.17, 14.33
Protective Circuits ........ 8.17, 12.22, 14.43, 14.44, 15.14, 15.16,
15.17,15.18,16.15,16.20,17.11
672
APPLICATION INDEX
Applications
For Basic Circuit Possibilities
See Figure Number
Pulse Modulators ......... 5.8, 5.9, 5.10, 5.11, 6.3, 6.10, 6.11, 6.12,
13.43
Regulated Power Supplies .. 6.3, 6.10, 6.11, 6.12, 6.25, 6.26, 6.27,
6.28, 9.40, 9.41, 9.42, 9.45, 9.46, 9.51,
9.52, 9.53, 11.7, 12.11
RFI Protection ........... 17.2, 17.3, 17.5
Ring Counters ........... 8.21, 8.22
Sensor Amplification ...... 8.19, 8.31, 8.32, 8.34, 8.35, 8.37, 9.30,
9.33,9.34,9.36,11.9, 1Ll1, 11.17, 12.6,
12.7, 12.10, 12.11, 12.19, 12.26
Static Relays. Contactors ... 4.20, 4.27, 6.25, 6.26, 6.27, 6.28, 7.7,
11.3, 11.4, 11.5, 11.6, 11.7; 11.8, 11.9
Temperature Controls ..... 6.3,6.10,7.7,8.33,11.3,11.4, 11.5, 11.6,
11.7,11.8, 11.9, 1Ll5, 11.16, 11.17, All
of Chapter 12
Thyratron Replacements ... 4.51, 6.3, 6.10; 6.11, 6.12, 8.38
Triggering Circuits ........ 4.10, 4.20, 4.21, 4.23, 4.25, 4.26, 4.27,
4.30; 4.32, 4.43, 4.44, 4.45, 4.46, 4.48,
4.49, 4.50, 4.51, 4.52, 4.53, 4.54, 6.10,
6.11, 6.12, 9.57, 9.58, 9.59, 9.60, 9.61,
9.62, 9.63, 9.64, 11.3, 11.4, 11.5, 11.6,
11.7, 11.8, 11.9, 11.10, 11.11, 11.12,
11.13, 12.6, 12.7, 12.11, 12.15, 12.18,
14.26, 14.27, 17.11
Timing Circuits .......... 4.30, 4.32, 8.23, 8.24, 8.25, 8.26, 8.28,
8.29, 8.30, 12.16, 14.35, 14.36
TransientVoltageProtection.6.9, 8.17,16.15,16.20,17.11
Ultrasonic Generators .... .4.12, 5.9, 5.20, 13.2, 13.32
673
SCR MANUAL
INDEX
AC Contactor,51
Battery Vehicle Controller, 369
AC Flasher Circuits, 209
Belleville Spring Washers, 519
AC Line Commutated, 352
Bending Leads, 511
AC Line Commutation, 397
Beryllium Oxide Washers, 507
Bias, Gate, 79, 81, 83, 87
AC Motor Control, 383
Bias, Gate, Positive, 79
AC Static Switches, 195
AC Time Delay Circuits, 216,
Bidirectional Thyristor, Diode, 183
. Bidirectional Thyristor, Triode, 181
219,220
Bilateral Switch, 183
Accessories - Inverter, 387
Bilateral Trigger Diode (Diac), 110
Active Stabilization, ,358
Blocking Characteristics, 567
Advantages - Solid State
Switching, 351
Blocking Current, 152
Air Flow Rate (Blower Driven), 544
Blocking Current, Peak Reverse, 567
Alloy-Diffused Structures, 15
Blocking Oscillator, 98
Alpha Analysis, Thyristor, 2
Blocking, Reverse, 4, 10
Alston 70 Process, 505
Blocking Voltage, 5, 10, 236-240
Blower Manufacturers, 545
Altitude EHects on Convection
Blower Motor Control, 343, 344,
Heat Transfer, 535
346,348
Amalgam - Fastening
Thermocouples, 550
Blower Selection, 544
Ambient Temperature, 606
Bombardment, Neutron, 559, 560
Amplification, Trigger Pulse, 119
Breakdown, Avalanche, 3
Amplifying Gate, 8
Breakover, Forward Voltage, 81
Analogy, Two Transistor, 1
Breakover Triggering, 182
Breakover Voltage, Forward, 11
Analysis of Phase Control, 232
Brightness Control, 444
Angular Response, 415
Brush Life, 289
Anode Characteristics, 71, 84
ANSI, 493, 500
Bum-In, 558
Apparent Thermal Resistance, 189
Applications - Inverter, 356,
Calibration - Force Gauge, 525
Candle Power, 422
383,399
Capacitance, Gate-Cathode, 77
Application Notes, 657
Capacitance, Heat Exchanger, 492
Arc Voltage, 452
Arrays, Triac, 191
Capacitance, Junction, 3
Capacitance, Stray, 492
Arrhenius Model, 560
Capacitive Load, 362
Assemblies, Triac, 344
Capacitor Boosting, 375
Asymmetric Triggering, 111
Capacitor, Current Limiting, 391
Asymmetrical Trigger Switch,
Capacitor, Snubber, 157, 158
183,111
Capacitor Start Motors, 299, 304
Audio Coded Inputs, 190
Capacitors, Commutation, 143
Auxiliary Commutated Inverter,
(Class D), 383
Capability, Thermal, 38
Case Mounting, Soldering, 509
Avalanche Breakdown, 3
Case Temperature, 606
Average Current, 565
Average Current Definition, 608
Case Temperature, Measurement, 549
Average Current Rating, 43
Catalogues, 657
Catastrophic Failures, 561
Back EMF, 289-294
Center Gate, 6
Characteristics, 23, 27
Back-Up Plates, 16
Characteristics, Anode, 71, 84
Battery Charger, 203
674
INDEX
Characteristics, Gate, 73, 88
Characteristics, Gate, Triac, 183
Characteristics, Voltage-Current,
Triac, 183
Charge Recovered, 69, 142, 157
Charge, Stored, 68, 495
Charger, Battery, 203
Charging - Resonant, 375
Check List, Selection, 600
Chopper Circuit, 133
Chopper Control, 370
Circuit Assembly
Specifications, 648-650
Circuit Design, Chopper, 378
Circuit Design, PWM Inverter, 385
Circuit Tum-Off Time, 127
Clamps, Press Pak, Packages, 523
Clamps, Press Pak, Thermal
Limitations, 526
Class A Inverter, 128, 354, 392
Class B Inverter, 130, 357
Class C Inverter, 131,361,392
Class D Inverter, 132, 383
Class E Inverter, 134
Class F Commutation, 397
Classification of Circuits, 128
Classification of Inverter
Circuits, 351, 352, 354
Classification, Thyristor, 1
Clearing Time, Fuse, 455
Clearing Time Nomograph, 455
Closed Loop Systems, 326
CoeffiCient, Temperature, UJT, 101
Color Temperature, 419, 420, 426
Commutatingdv/dt,66
Commutating dv I dt test, 588
Commutation AC Line
Commutated, 397
Commutation Capacitors, 143
Commutation Circuit Design, 385
Commutation, Class D, 383
Commutation Classes, 352, 354
Commutation dv I dt, 187
Commutation - LC Switched, 352
Commutation Methods, 127,
128,352
Commutation, Phase Control, 246
Commutation Properties, 352, 353
Commutation - Self, 352
Comparison of Power
Semiconductors, 20
Compensation, Line Voltage,
261,262
Complementary SCR, Definition, 4
Components, Filter, 493, 494
Concurrent dil dt Rating, 54
Concurrent High Frequency
Ratings, 57
Conducted Interference, 489
ConductiVity - Thermal, 538
Construction, Photo SCR, 418
Contactor, AC, 51
Contacts, Pressure, 18
Contaminants, Internal, 558
Control, Cooling, 346, 348
Control, Feedback, 327
Control, Heater, 325
Control, On-Off, 335
Control, Proportional, 327, 328, 337
Control, Remote, 338
Control, Temperature, 223, 300-302,
317,325,346
Convection -Forced Heat
Transfer, 535
Convection - Heat Transfer, 533
Cooling Control, 346, 348
Cooling - Lead Mounted, 503
Co-ordination Chart, 463
Core Design, 379
Corrosion Inhibitors, 505, 507
Counter EMF, 289-294
Counters, Ring, 213, 214
Coupler, Photo, 425, 426
Critical Rate of Rise of On-State
Current Test, 582
Critical Rate of Rise of Off-State
Voltage Test, 585
Crowbar, 486
Current, Average, 565, 608
Current, Average Rating, 43
Current, Blocking; 152
Current Derating, Parallel
Operation, 174
Current, Fault, 447
Current, Form Factor, 241-244
Current, Forward, 2
Current, Free Wheel Diode, 236-240
Current, Gate, 2, 4, 10
Current, Gate, Negative, 75, 81
Current, Gate Trigger, Test, 570
Current Holding, 3, 30, 67, 77, 89
Current, Holding, Test, 576
Current, Holding, Tester, 577
Current, Inrush, 245, 421, 423
Current in SCR, 236-240
675
SCR MANUAL
Current, Latching, 30, 67, 89,316
Current, Latching, Test, 578
Current, Latching, Tester, 578
Current, Latching, Triac, 199
Current, Let-Thru, 451
Current, Limit, Pulse Width
Control, 389
Current Limiter, 444
Current Limiting, 451
Current Limiting Fuse, 451
Current, Line, 232, 234
Current Measurement, 602
Current, Non-Recurrent Rating,
42,45
Current, Peak OH-State, 567
Current, Peak Point, 99
Current, Peak Reverse Blocking, 567
Current Probe, 602
Current, Prospective Fault, 452
Current Protective Circuits, 210, 212
Current Rating, Fault, 42, 45
Current Rating - High
Frequency,56
Current Rating, Multicycie, 45, 46
Current Rating, Phase Control, 43
Current Rating, Power Tab, 43
Current Rating, Press Pak, 43
Current Rating, RMS, 44
Current Rating, Subcycie, 46
Current Rating, Welding, 49
Current Ratings, 42
Current Ratings - Rectangular, 59
Current Ratings, Sinusoidal Wave
Shape,56
Current Regulator, 262
Current, Reverse Recovery, 68
Current, RMS, 608
Current Sensing Cu:cuit, 221
Current Sharing, 171, 176
Current Shunt, 603
Current, Surge, 30
Curve Tracer to Measure
Thyristors, 595
Cycioconverters, 397
Cycioinverter, 396
Darlington Amplifier, Light
Activated, 413
DC Flasher Circuits, 205, 206,
207,208
DC Loads, 272-279
DC Motor Control, 370
676
DC Static Switches, .204
DC Time Delay Circuits, 215,217,
218,219
DC Triggering, 85
Definition, SCR, 1
Delay Circuits, AC, 216, 2,19, 220
Delay Circuits, DC, 215, 217,
218,219
Delay Reactor, 141
Delay Time, 6, 8, 32, 34, 90,
151,167
Density, Power, 52, 57
Department of Defense, 500
Derating, 560
Derating, EHects of, 560
Design,Chopper Control, 378
Design, Commutation Circuit, 385
Design, Filter, 365
Design, Flat Fin Heat
Exchanger, 530
Design; Inverter Class C, 366
Design, PWM Inverter, 385
Design Trade-Off's; 599
Detector, Light, 440
Detector, Proximity,224
Detector Specifications, 638
Detector, Threshold, 224
Diac, 25, 110, 183, 192
di/dt, 5, 6,10,75,78,91,119,
141,604
dil dt, Recurrent, 53
di/dt, Test, 582
di/dt, Tester, 582
di/dt, V(BO) Triggering, 55
Diffused Pellet Construction, 14
Diode, Feedback, 127
Diode, Free-Wheeling, 236-240,·
371,379,486
Diode, Light Emitting, 423
Diode, Photo, 409
Diode, Trigger, 110
Diode, Tunnel, 23
Discharge, Resonate, 492
Distributor Sales Offices, 666
Driver, Neon Tube, 228
Duty Cycie, 59
dv/dt,3,9,29,83,497,604
dv/dt, Commutating, 66, 187
dv/dt Rating, 63
dv/dt, Reapplied, 63,140
dv I dt, Static, 64
dv I dt, Suppression, 139
INDEX
dv/dt,
dv/dt,
dv/dt,
dv/dt,
dv/dt,
Tes~,
585
Test, Exponential, 586
Test, Linear, 586
Tester, Exponential, 586
Tester, Linear, 587
Effective Irradiance to Trigger, 429
Effective Thermal Resistance, 188
Effects, Ground, 490
Effects of Derating, 560
Efficiency, Luminous, 430
Efficiency, Quantum, 424
Egg Crate Curves, 455
Electric Vehicle Controller, 369
Electrical Isolation
Case to Heat Exchanger, 507
Electrical Isolation Using
Mylar Tape, 509
Electrical Isolation, Press
Fit Packages, 508
Electrical Solution, Stud
Packages, 508
Electromagnetic Interference, 489,
Electromagnetic Interference
Standards, 500
Electronic Flash, 441
Elevated Temperature Testing, 598
Emergency Lighting System, 226
EMF, 289-294
EMI,307
Emissivity, Surface, 532
Emittance, Tungsten Lamps, 409
Emitter Shorts, 9, 73
Encapsulation, 16
Encapsulation Flaws, 558
Encapsulation, Plastic, 19
Equalizing Network, 150
Equivalent Circuit, Gate, 73
Equivalent Circuit, Thermal, 38
Example Design, Chopper
Circuit, 380
Example Design, PWM Inverter, 386
Exponential dv/dt, Test, 586
External Pulse Commutation, 352
Factor, Form, 44, 241-244, 566,609
Failure Mechanisms, 5, 6, 7, 557,
558,559
Failure Rate, 554, 555, 560, 561, 563
Failures, Catastrophic, 561
Fall Time, 90
False Triggering, 497
Fan Motor Control, 300
Fast Recovery Rectifiers, 495
Fatigue, 555
Fatigue, Thermal, 16
Fault Current, 447
Fault Current Rating, 42, 45
FCC, 500
Federal Communications
Commission, 489, 500
Feedback Control, 327
Feedback Diode, 127,359,361
F.1. Gate, 6
Field Initiated Gate, 6
Filament Temperature, 419
Filter, Harmonics, 395
Filter Components, 494
Filter Design, 365
Filter, LC, 392
Filter, Ott, 362
Filter, RFI, 490, 493
Filter, Switching, 393
Filtering, RFI, 490
Fin, Convection, 533
Fin Design, Worked Example, 539
Fin Effectiveness, 536
Fin Forced Convection Heat
Transfer, 535
Fin, Radiation, 532
Flash, Electronic, 441
Flasher, 191
Flasher Circuits, AC, 209
Flasher Circuits, DC, 205, 206,
207,208
Flat Base Package Mounting, 519
Flat Fin Heat Exchanget"
Design, 530
Flip Flop, 118, 204, 205
Flux Density, 379
Force Gauge, Press Pak Clamp, 525
Forced Commutation, 128
Forced Convection Trade Offs, 543
Forced Convection Heat
Transfer, 53.5
Forced Cooling Design, 544
Forced Current Sharing, 176
Form Factor, 44, 241-244, 566, 609
Forward Breakover Voltage, 11
Forward Breakover Voltage, 81
Forward Characteristics,
Matched, 173
Forward Current, 2
Free Wheeling Diode, 236-240, 371,
379,486
677
SCR MANUAL
Frequency Distribution of SCR, 490
Fullwave Motors, 291
Fuse, Application, 453
Fuse, Arc Voltage, 452
Fuse, Clearing Time, 455
Fuse, Current Limiting, 451
Fuse Ratings, 453
Fuse, SCR Application Chart, 455
Fuse, SCR Coordination, 451
Fusing, Inverters, 389
G-4,175
G-7,175
Gain, Turn-Off, 12
Gamma Radiation, 8, 559, 560
Gas-Filled Lamps, 422
Gate, Amplifying, 8
Gate Bias, Positive, 79
Gate Bias, 79, 81, 83, 87
Gate Blocking Protection, 466
Gate-Cathode Capacitance, 77
Gate-Cathode Equivalent Circuit, 73
Gate Cathode Impedance, 78
Gate Cathode Inductance, 78
Gate-Cathode LC Resonant
Circuit, 79
Gate Cathode Resistance, 76
Gate, Center, 6
Gate Characteristics, 73, 88, 166
Gate Circuit Damage, 81, 85, 86
Gate Connection, Parallel, 116
Gate Current, 10
Gate Currents, Negative, 75, 81
Gate, Equivalent Circuit, 73
Gate, Field Initiated, 6
Gate, Interdigitated, 9
Gate Jnnction, 83, 182, 184
Gate Losses, 37
Gate, N+, 8
Gate, Point, 6
Gate Power, 85, 86, 87
Gate, Remote, 182, 185
Gate, Side, 6
Gate Source Impedance, 76
Gate Structures, 6
Gate Test, Anode Supply, 571
Gate Test Circuits, 572-575
Gate Test, Curve Tracer, 596
Gate Test, Gate Supply DC, 572
Gate Test, Gate Supply Pulsed, 573
Gate Test, Low Current SCR's, 574
Gate Transients, 499
678
Gate Trigger Characteristics,
Triacs, 183
Gate Trigger Current, 2, 4, 10
Gate Trigger Current, Test, 570
Gate Trigger Voltage, Test, 570
Gate Triggering, 71, 85
Gate Turn"Off Switch, 12
Gauge, Clamp Force, 525
GE-MOV, 159, 477,481
GE-MOV Specification, 656
Generators, Trigger Pulse, 98
Glass Passivation, 20
Graphical Symbols, 23-25
Grease, Thermal, 505
Ground Effects, 490
Guggi Circuit, 358
Half Wave Motors, 288
Handling of Press Paks, 529
Hard Solder, 16
Hardware Kits, 507
Hardware Kits, Insulated, 512
Harmonic Reduction~ 395
Head Loss, 544
Heater Control, 325
Heat Exchanger, Blower .'
Selection, 544
Heat Exchanger, Finish, 505
Heat Exchanger, Flat Fin, 530
Heat Exchanger, Flatness, 505
Heat Exchanger, Manufacturers, 545
Heat Exchanger Selection, 530
Heat Exchanger Selection
Guide, 542
Heat Exchanger Selection,
Liquid Cooled, 546
Heat Exchanger, Smoothness, 505
Heat Exchanger, Surface
Preparation, 505
Heat Exchanger, Time Constant, 41
Heat Exchanger, Volume
Requirements, 542
Heat Exchangers, Commercial, 541
Heatsink, 16, 613
Heatsink, Capacitance, 492
Heatsink, Transient Effects, 41
Heat Transfer, Convection, 533
Heat Transfer, Forced
Convectiol)., 535
Heat Transfer, Liquid Cooled, 546
Heat Transfer, Radiation, 532
High Frequency Current Ratings, 56
INDEX
High Frequency Voltage Ratings, 63
High Voltage Switch, 437
Holding Current, 3, 30, 67, 77, 89
Holding Current Test, 576
Holding Current Tester, 577
Hole Storage, 5, 10
Impedance, Capacitor, 494
Impedance, Gate-Cathode, 78
..Impedance, Gate, Source, 76
Impedance, Inductor, 494
Incandescent Lamps, 245
Indicator, Transient Voltage, 475
·Indirect Feedback, 331
Inductance, Gate Cathode, 78
Induction Motors, 287, 298-305
Inductive Kick, 247
Inductive Loads, 241-244, 265
Input Filter Impedance, 363
Inquiries, 657
Inrush Current, 245, 421, 423
Insulated Hardware Kits, 512
Integrated Phase Control, 267, 303
Intensity, Light, 426, 427
Interaction, 497
Interbase Resistance, 101
Interdigitated Gate, 9
Interface, Low Level Logic, 191
Interface, Thermal, 504, 506
Interface, Thermal Grease, 505, 507
Interface Thermal Resistance,
Power Pac, 516
Interface Thermal Resistance
for Power Tab SCR's, 513
Interference, 489
Interference, Radiated, 496
Interference with Radio & TV, 307
Internal Contaminants, 558
International Sales Offices, 666
Inverter Applications, 356, 383
Inverter, Auxiliary Commutated, 383
Inverter Circuit Configurations, 353
Inverter Circuit Definition, 351
Inverter, Class A, 128, 352, 354, 392
Inverter, Class B, 130, 352, 357
Inverter, Class C, 131, 352, 361, 392
Inverter, Class D, 132, 352
Inverter; Class E, 134,352
Inverter, Class F, 352
Inverter Class. Properties, 352
Inverter Commutation Methods, 128
Inverter Design, Class C, 366
Inverter, Guggi Circuit, 358
Inverter, McMurray Bedford, 361
Inverter, Overcurrent Protection, 389
Inverte~ PVVM,383
Inverter, Reactive Load
Operation, 388
Inverter, RFI, 489
Inverter, Sinewave Output, 355, 357,
362,383,392
Inverter Trigger Circuits, 118
Inverter, UPS, 383
Irradiance, 427
Irradiance Calculations, 430
lrradiance, Effective to Trigger, 429
Isolation, Electrical, Case to
Heat Exchanger, 507
Isolation, Thermal, Press Fit
Packages, 508
Isolation, Thermal, Stud
Packages, 508
Isolation Using Mylar Tape, 509
I Squared t (I2 t), 32
I 2 t Ratings, 45, 46
Jones Chopper Circuit, 133, 369
Junction Capacitance, 3
Junction, Gate, 83, 182, 184
Junction Temperature, 35
Kick, Inductive, 247
Lambda, 554, 556,560
Lamp, Inrush, 245
Lamp, Solid State, 423
Lamps, 245
Lamps, Gas Filled, 422
Lamps, Tungsten, 419-423, 426
LASCR, 414, 417
LASCS,418
Laser Pulser, 397
Latching Current, 30, 67, 89, 316
Latching Current Test, 578
Latching Current Tester, 578
Latching Current, Triac, 199
Lateral Resistance, 9, 73
Layout,. 499
Layout, VViring, 490
LC Filter, 392
LC Resonance, Current Limit, 390
LC Switched Commutation, 352
Lead Bending, 511
Lead Configuration, Power Pac, 516
679
SCR MANUAL ,
Lead Configuration, Power Tab, 512
Lead Mounted Device Cooling, 503
LED Specifications, 640, 641
Let-Thru Current, 451
Level Control, 226
Light Activated Darlington
Amplifier, 413
Light Activated High Power
SCR's. 433, 434, 437 .
Light Activated High Voltage
Switch, 437
Light Activated Logic Circuits, 438
Light Activated Motor Control, 442
Light Activated Phase Control, 436
Light Activated Relay, 432, 435
Light Activated SCR, 4
Light Activated Silicon Controlled
Switch (LASCS), 418
Light Activated Thyristor,
LASCR,414
Light Activated Triac, 434, 435
Light Activated Zero Voltage
Switch,435
Light Control, 264
Light Detector, 440
Light Emitting Devices, 419
Light Emitting Diode, 423
Light Intensity, 426, 427
Light Sensing Circuits, 439, 440, 441
Light Sensitive Transistor, 411
Light Sensitivity, 416
Light Sources, 431
Light Triggering, 161, 162
Light Triggering Characteristics, 415
Lighting; 245
Lighting System, Emergency, 226
Lightning Transient, 470
Limiter, Current, 444
Line Voltage Compensation,
261,262
Linear dv/dt Test, 586
Linear Phase Control, 333
Liquid Cooling, 546
Liquid Level Control, 226
Liquid Selection for Heat
Exchangers, 548
Load Current, 232, 234
Load Impedance, 363
Load Power Factor Effect, 355, 359
Load Regulation, 391
Load Voltage, 232, 234, 236-240
Loads, Inductive, 241-244
680
Locked Rotor, 378
Logic, Low Level, Interface,c 191
Logic Circuits, Light.Activated, 438
Losses, Gate, 37
Losses, High Frequency, 5&-60
Losses, On-State"37
Losses, Switching, 56, 57
Lot Tolerance Percent
Defective,554
Low Level Logic Interface,c191 L TPD, 4, 554, 556
Lubricant, Thermal Grease, 505, 507
Luminous Efficiency, 430
Magnetic Amplifier Trigger
Circuit, 96
Matched Characteristics, 173
Material, Thermal Properties, 538
McMurray Bedford Inverter,.
332,361
Mean Time Between Failures, 554
Measuring Transients, 473
Measurement, Case
Temperature, 549
Measurement, Current, 602
Measurement, Temperature, 606
Measurement, Temperature of
Power Pack Package, 514
Measurement, Voltage, 602
Mechanism, Tum-Off, 4
Mechanisms, Failure, 557
Metal Oxide Varistor, 159,477,481
481,656
Meter; Peak Recording, 474
Mica Washers, 507
Modes of Triggering, Triac, 183-184
Monitor, Temperatures, 222
Morgan Circuit, 130, 352
Motor Control, AC, 383
Motor Control, DC, 370
Motor Control, Light Activated, 442
Motors, Back EMF, 2&7, 289-294
Motors, Brush Life, 289
Motors, Capacitor Start, 299, 304
Motors, Fan Control, 300
Motors, Fullwave, 291
Motors, Halfwave, 288
Motors, Induction, 287, 298-305
Motors, Induction Starter, 304-305
Motors, Phase Control, 287
Motors, PM, 292
Motors, Reversing, 297, 303-304
INDEX
Motors, Rotor Resistance, 300
Motors, Series, 288
Motors, Shunt Wound, 292
Motors, Speed Control, 287
Motors, Split Phase, 299
Motors, Universal, 287
Mounting Clamps for Press Pak
Packages, 523
Mounting, Flat Base Packages, 519
Mounting, Interface
Resistance, 504, 506
Mounting, Multiple Unit, 527
Mounting, Power Pac
Package, 514, 515
Mounting, Power Tab, 511
Mounting, Press Fit Package, 517
Mounting, Press Pak Package, 522
Mounting, Stud Packages, 518
Mounting, Temperature Cycling, 519
Mounting, to Heat Exchangers, 504
Mounting, Torque, 519
Mounting, Unit Pak Package, 529
MTBF,554
Multicycle Current Rating, 45, 46
Mylar Tape, Isolation, 509
N+ Gate, 8
Negative Gate Currents, 75, 81
Negative Pulse Triggering, 94
Negative Resistance, 71, 98, 100
NEMA, 493, 500
Neon Lamp Trigger, 114
Neon Trigger, 296
Neon Tube Driver, 228
Neutron Bombardment, 8, 559, 560
Noise, Electrical, 490
Noise, SCR, 490
Non-Recurrent Current Rating,
42,45
Non-Repetitive Voltage, 29
Nuclear Radiation, 559, 560
Off-State, 4, 60
Off-State Test on Curve Tracer, 596
Offices, Sales, 660
Oil, Thermal Lubricant, 506, 507
On-Off Control, 335
On-State, 2, 28
On-State Losses, 37
On-State, Voltage Test,
High Level, 580
On-State, Voltage Test,
Low Level, 579
On-State, Voltage Test
on Curve Tracer, 597
On-State Voltage Tester, 581
"One Shot" SCR Trigger Circuit, 202
OpticwTriggering, 161, 162
Optoelectronic Specifications,
638-641
Oscillator, Blocking, 98
Oscillator, Relaxation, 98
Oscillators, 98
Ott Filter, 362
Output Voltage Switching, 393
Overcurrent Protection for
Inverters, 389
Overload Protection, 447
PA436,303
Package Configurations,
Power Tab, 512
Package, Plastic, 19
Parallel Gate Connection, 116
ParwlelOperation, 149, 165
Parallel Mounting, Press Paks, 527
Parallel Triggering, 178
Passivation, Glass, 20
Passivation, Planar, 15
Passive Stabilization, 358, 361
Peak Off-State Current Test, 567
Peak Off-State Voltage Test, 567
Peak On-State Voltage Test, 579
Peak Point Voltage or Current, 99
Peak Reading Volbneter, 569
Peak Recording Meter, 474
Peak Reverse Blocking Current, 567
Peak Reverse Voltage, 236-240, 567
Peak Value, 567
Pellet, Diffused Construction, 14
Pellet, Fabrication, 14
Pellet, Structure, Triac, 182
Permanent Magnet Motors, 292
PFV Rating, 61
Phase Control, 91,95, 114, 117,
231,331
Phase Angle, Maximum Cutoff, 278
Phase Control, Current Rating, 43
Phase Control, DC Loads, 272-279
Phase Control, Frequency
Selective, 260
Phase Control, Inductive
Load,241,265
Phase Control, Integrated
Circuit, 2m, 282
681
SCR MANUAL
Phase Control, Light Activated, 436
Phase Control, Motors, 287-305
Phase Control, Negative Ramp, 267
Phase Control, Pedestal, 256
Phase Control, Polyphase, 276
Phase Control, Ramp, 254
Phase Control, Ramp & Pedestal, 257
Phase Control, RC, 91
Phase Control, Triac, 252
Phase Control, Trigger Circuit, 249
Phase Shift Trigger Circuits, 95
Phase, Three, 276
Photo Darlington Amplifier, 413
Photo Diode, 409
Photo SCR Construction, 418
Photo Thyristor (LASCR), 414
Photo Transistor, 411
Photon Coupler, 425, 426
Pilot SCR, 6, 74, 120, 272
Planar Passivation, 15
Plastic Encapsulation, 19
Plates, Back-Up, 16
Plating, Nickel or Cadmium, 505
PM Motors, 292
Point Gate, 6
Polyphase Application, 276, 399
Polyphase Phase Control, 276
Positive Gate Bias, 79
Post, Liquid Cooled, 547
Power Density, 52, 57
Power Dissipation, 36, 155
Power Dissipation,
High Frequency, 58-60
PowerFacto~Load,355,359,387
Power, Gate, 85, 86
Power Pac Package, Interface
Thermal Resistance, 516
Power Pac, Lead Configuration, 516
Power Pac, Package
Mounting, 514, 515
Power Pac, TO-66 Mounting,
515,516
Power Tab, Current Rating, 43
Power Tab, Lead Bending, 511
Power Tab Mounting, 511
Power Tab Package
Conflgurations, 512
Press Fit Insolation, 508
Press Fit Package Mounting, 517
Press Pak, 18
Press Pak, Clamp, Thermal
Limitations, 526
682
Press Pak, Clamping
Requirements, 523
Press Pak, Current Rating, 43
Press Pak, Handling, 529
Press Pak, Mounting Clamp, 524
Press Pak, Package Mounting, 522
Press Pak, Parallel Mounting, 527
Press Pak, -Series Mounting, 528
Press Pak Vs Stud, 542
Pressure Contacts, 18
Pressure Drop, 544
Pressure Switches, 190
Probe, Current, 602
Programmable UJT, 25, 105
Properties, Water, 549
Proportional Control, 327, 328, 337
Prospective Fault Current, 452
Protection, Co-ordination, 449
Protection, Gate Blocking, 466
Protection, Overcurrent, 447
Protection, Redundancy, 464
Protection, Surface, 558
Protection, Systems, 447
Protective Circuits,· Current,
210,212
Protective Circuits, Voltage, 209
Proximity Detector, 224
Pulse Amplification, 119
Pulse Circuits for Light
Emitters, 445, 446
Pulse Modulator Switches, 397
Pulse Tranformers, 115
Pulse Triggering, 81
Pulser SCR, 391
PUT,25,105
PUT Specifications, 636
PWM Inverter, 383
Quality,553
Quantum Efficiency, 424
Radar Modulators, 391
Radiated Interference, 490, 496
Radiation, Gamma, 559, 560
Radiation, Nuclear, 559, 560
Radiation, Thermal, 532
Radiation Tolerance, 559, 560
Radio Frequency Interference, 489
Radio Frequency Interf~rence .
Standards, 500
Ramp & Pedestal, 257, 333
Ratcheting, Thermal, 519
INDEX
Rate of Rise of Current, 141
Rate of Rise of Voltage, 139
Ratings, 27, 42
Ratings, Current, 42
Ratings, dil dt, 52
Ratings, dil dt, Concurrent, 54
Ratings, dv I dt, 63
Ratings, Fuse, 453
Ratings, High Frequency, 56
Ratings, I2t, 45, 46
Ratings, Inductive Loads, 241-244
Ratings, PFV, 61
Ratings, Rectangular, High
Frequency, 58
Ratings, References, 69
Ratings, Subcycle, 46
Ratings, Surge, 45, 46
Ratings, Voltage, 60
Ratings, Voltage
( High Frequency), 63
RC Phase Control, 91
Reactive Load, 388
Reactor, Delay, 141
Reactor, Saturable, 141
Reapplied dv/dt, 29, 65, 140
Recombination, 10
Recovered Charge, 69, 142, 155
Recovery Characteristics, 68
Recovery Current, 68
Recovery, Reverse, 142,495
Recovery Time, 32, 68
Rectangular, High Frequency
Ratings, 58
Rectifier Diode, Free
VVheeling, 371, 379
Rectifier, Fast Recovery, 495
Rectifie~Feedback,359,361
Rectifier, Reverse Recovery, 495
Rectifier, RFI, 495
Rectifier & SCR Module
Specification, 651-653
Rectifier Snap Off, 495
Rectifier Specifications, 642-647
Redundancy, for Fault
Protection, 464
Reed Switch, 190
Reference Point, Thermal for
Power Tab, 512
References, Ratings, 69
Regulation, Load, 391, 355, 359,
383,387
Regulator, Current, 262
Regulator, Light, 264
Regulator, Voltage, 252
Relaxation of Studs, 519
Relaxation Oscillator, 98
Relay, Light Activated, 432-435
Relay, Resonant Reed, 190
Relay, Solid State, 196,200
Reliability, 553, 554, 555, 563
Reliability Screens, 563
Remote Base Transistor, 13,228,318
Remote Control, 338
Remote Gate, 182, 185
Resistance, Apparent Thermal, 189
Resistance, Effective Thermal, 188
Resistance, Gate-Cathode, 76
Resistance, Interbase, 101
Resistance, Lateral, 9, 73
Resistance, Negative, 71, 98, 100
Resistance, Rotor, 300
Resistance, Shunt, 152
Resistance, Thermal, 31, 34, 37
Resistance, Thermal, Press Pak, 595
Resistance Thermal, Test, 591
Resistance, Thermal, Tester, 594
Resistance, Triac Thermal, 188
Resonant Charging, 375
Resonant Circuit, Gate Cathode, 79
Resonant Discharge, 492
Resonant Loads, 392
Resonant-Reed Relays, 190
Response, Angular, 415
Response, Spectral, 416
Reverse Blocking, 4,10
Reverse Recovery, 142
Reverse Recovery Characteristics, 68
Reverse Recovery Current, 68
Reverse Recovery Time, 32
Reverse Voltage, 61
Reverse Voltage Test on
Curve Tracer, 596
Reversing Drives, 297
Reversing, Motor, 303, 304
RFI,307
Ring Counters, 213, 214
Rise Time, 90
RMS Current Definition, 608
RMS Current Rating, 44
RMS Value, 565
Rotor Resistance, 300
Sales, Distributor, 660
Sales Offices, 660
683
SCR MANUAL
Sample Sizes, 554, 555, 556, 560
Sampling Plan, 554, 556
Saturable Reactor, 141
Saturable Reactor, Trigger
Circuits, 96
SBS, 110, 183
SCR Selector Chart, 616, 623
SCR Specifications, 617, 628
SCR Stack Assembiies, 599
SCR, Turn-On, 492
Screens, Reliability, 563
SCS, 4,113
Selecting a Heat Exchanger, 530
Selection Check List, 600
Selection Guide, Heat
Exchangers, 542
Selection, Thyristor, 599
Selector Chart, SCR, 616, 623
Selenium Components, 654-655
Self Commutated Inverter, 352
Semiconductor Trigger-Pulse
Generators, 98
Sensing Circuit, Current, 221
Sensing Circuit, Voltage, 224
Sensing Circuits, Light, 439-441
Sensitivity, Light, 416
Series Motors, 288
Series Mounting Press Paks, 528
Series Operation, 149
Servo, Reversing Drive, 297
Sharing, Current, 171, 176
Sharing, Voltage, 159
Shielding, 491, 496
Shielding Materials, 496
Short Circuits, 447
Shorted Emitter, 9
Shorts, Emitter, 9, 73
Shunt, Current, 603
Shunt Resistance, 152
Shunt Wound. Motors, 292
Side Gate, 6
Silicon Bilateral Switch, 110, 183
Silicon Controlled Switch, 26, 113
Silicon Unilateral Switch, 109
Sine Wave Inverter, 355, 357,
383, 392
Sink, Heat, 16, 613
Sinusoidal Wave Shape Current
Ratings, 56
Slave Triggering, 94,163,200
Snap-Off, Rectifier, 495
Snap-On, 252
684
Snubber, 151, 157
Snubber Capacitor, 157, 158
Snubber Circuits, 83, 139, 481
Snubber, Triac, 187
Soft Solder Construction, 18
Soft Start, 259, 260
Soft Stop, 259
Solder, Case Mounting, 509
Solder, Hard, 16
Solder, Soft, 18
Solid State Lamp, 423
Solid State Lamp
Specifications, 640-641
Solid State Relay, 196, 200
Sources, Light, 431
Specifications, Circuit
Assemblies, 648-650
SpeCifications, GE-MOV, 656
Specifications, LEDS, 640-641
Specifications, Optoelectronics,
638-641
Specifications, PUT, 636
Specifications, Rectifiers, 642-647
Specifications, Rectifier & SCR
Modules, 651-653
Specifications, SCR, 617-628
Specifications, Switches; 636-637
Specifications, Thyrectors, 655
SpeCifications, Triac, 629-632
SpeCifications, Triac Triggers,
633-637
Specifications, Unijunctions, 634-636
Spectral Distributions of LED &
Light Sensitive Devices, 427
Spectral Response, 416
Speed, Switching, Light Sensitive,
Devices, 412
Split Phase Motors, 299
Spreading, Heat, 509, 510
Spring Washers, 519
Stability, 588
Stability, Thermal, 555
Stabilization, Active & Passive, 358
Standards, RFI, 500
Start Switch, 304-305
Saturation of SCR's, 72
Static AC Switches, 195"
Static DC Switches, 204·
Static dv/dt, 64.
Static Switching, 189
Storage, Hole, 5, 10
Storage Temperature, 35
j,
INDEX
Stored Charge, 68, 495
Structural Flaws, 557
Structure, Pellet, Triac, 182
Stud Isolation, 508
Stud Package Mounting, 518
Stud Torque, 519
Subcycle Current Rating, 46
SubSCripts, 27
Supply Voltage Regulation, 396
Suppresison Techniques, 477
Suppression, Voltage, 477
Surface Preparation, 505
Surface Protection, 558
Surface Thennal Emissivity, 533
Surge Current, 30
Surge Ratings, 45, 46
SUS, 109
Switch, Gate Turn-Off, 12
Switch, Reed, 190
Switch, Start, 304-305
Switch, Touch, 224
Switches, AC Static, 195
Switches, DC Static, 204
Switches, Pressure, 190
Switches, Specifications, 636-637
Switches, Timer, 190
Switching Interval, 40
Switching Losses, 56, 57
Switching Speed, Light Sensitive
Devices, 112
Switching Static, 189
Switching, Zero Voltage, 201,
307,334
Symbols, 33
Symbols, Graphical, 23-25
Syncronization, Relaxation,
Oscillators, 117
Synchronous Switching, 201, 307
Tachometer, 302-303
Temperature, Case, 606
Temperature Coefficient, UJT, 101
Temperature, Color, 419, 420, 426
Temperature Control, 223, 300-302,
317,325
Temperature Cycling Effects, 519
Temperature, Elevated, Testing, 598
Temperature, Filament, 419
Temperature, Junction, 35
Temperature Measurement, 549, 606
Temperature Measurement, Power
Pac, 514
Temperature Measurement,
Power Tab, 512
Temperature Monitor, 222
Temperature Rise, 37
Temperature, Storage, 35
Test Circuits for Thyristors, 565-598
Theory of Operation, I
Thennal Capacity, 38
Thermal Circuit, 38
Thennal Conductivity, 538
Thennal Considerations,
Press Pak Clamping, 526
Thermal Fatigue, 16
Thennal Grease, 505, 507
Thennal Impedance, Transient, 37,
38,39,40,540
Thennal Properties of Materials, 538
Thermal Radiation, 532
Thennal Ratcheting, 519
Thennal Resistance, 31, 34, 37, 612
Thermal Resistance, Apparent, 189
Thermal Resistance, Case to
Heat Exchanger, 504, 506
Thermal Resistance, Effective, 188
Thermal Resistance, Interface
For Power Tab SCR's, 513
Thennal Resistance Test, 591
Thennal Resistance Test,
Press Pak, 595
Thennal Resistance Tester, 594
Thennal Resistance, Triac, 189
Thermal Stability, 555
Thennal Lubricant, Oil, 506, 507
Thennistor Control Circuits, 329
Thennocouples, Used for Case
Temperature Measurement, 540
Thennostat, 190
Thennostat Control, 346
Threshold Detector, 224
Thyratron Replacement, 227
Thyrector, 480
Thyrector Specifications, 655
Thyristor Alpha Analysis, 2
Thyristor, Bidirectional Triode, 181
Thyristor Classification, 1
Thyristor Definition,. 1
Thyristor Design Trade-Offs, 5
Thyristor Triggering; 3
Thyristor Selection, 599
Thyristor Two Transistor Analogy, 1
Time, Clearing, 455
Time Constant, Heat Exchanger, 41
685
SCR MANUAL
Time, Delay, 6, 8, 32, 34, 90,151, ]67
Time Delay Circuits,
AC, 216, 219, 220
Time Delay Circuits,
DC 215,217,218,219
Time, Fall, 90
Time, Recovery, 68
Time, Reverse Recovery, 32
Time, Rise, 90
Time, Tum-Off, 5, 10, 124, 247, 248
Time, Tum-Off, Circuit, 127
Time, Tum-Off Test, 589
Time, Tum-Off, Tester, 590
Time, Turn-On, Gate Controlled, 584
.Time, Tum-On, Tester, 584
Timer Switches, 190
TO-66 Package, Power Pac
-Equivalent, 515; 516
TO-220 Package, See Power Pac, 514
Tolerance, Radiation, 559-600
Torque, Stud, 519
T-ouch Switch, 224
Trade-Off's in Thyistor Design, 5
Trade-Off's, Design, 599
Transformer, Voltage Transient, 485
Transformers, Pulse, 115
Transient Effects, Heat Sink, 41
Transient, Gate, 499
Transient Thermal Impedance, 37,
38,39,40,540
Transient, Trigger Circuit, 498
Transient Voltage Indicator, 475
Transients, 489
Transients, Voltage, 469
Transistor Action in SCR, 1
Transistor, Darlington, Light
Activated, 413
Transistor, Light Sensitive, 411
Transistor, Photo, 411
Transistor, Remote Base, 13,228,318
Triac, 26, 181
Triac Arrays, 191
Triac Assemblies, 344
Triac Circuits, 192
Triac, Commutation, 186
Triac, Commutation dv / dt, 66
Triac, Light Activated, 434, 435
Triac Pellet Structure, 182
Triac Specifications, 629-632
Triac, Speed Control, 493
Triac, Theory, 184
Triac, Thermal Resistance, 188
686
Triac Trigger Specifications, 633-637
Triac Triggering, 314
Triac, Tum-On, 492
Trigger Angles, Maximum
& Minimum, 249
Trigger, Asymmetrical, 183
Trigger Characteristics, Triac, 183
Trigger Circuit, ..Mag Amp, 96
Trigger Circuit, SCR One-Shot, 202
Trigger Circuits, 85, 88, 91
Trigger Circuits, di/dt Test, 584
Trigger Circuits, Interaction, 498
Trigger Circuits, Inverter, 118
Trigger. Circuits, Phase Control, 249
Trigger Circuits, Phase Shift, 95
Trigger Circuits, Saturable
Reactor, 96
Trigger Diode, Bilateral, 110
Trigger, Light, 161, 162
Trigger, Neon, 296
Trigger, Neon Lamp, 114
Trigger Pulse Amplification, 119
Trigger Pulse Generators, 98
Triggering, 3, 604
Triggering, Asymmetrical, III
Triggering, Breakover, 182
Triggering Characteristics, Light, 415
Triggering, DC, 85
Triggering, False, 497
Triggering, Gate, 71, 85
Triggering Modes of Triacs, 183, 184
Triggering, Negative Pulse, 94
Triggering, Optical, 161, 162
Triggering, Parallel, 178
Triggering Process, 7I
Triggering, Pulse, 87
Triggering, Simultaneous, 160
Triggering, Slave, 94, 163, 200
Triggering Triacs, 183,314
Triggering, V(BO), di/dt
Capability, 55
Tubing Manufacturers, 549
Tungsten Lamp Emittance, 409
Tungsten Lamps, 419-423, 426
Tunnel Diode, 23
Tum-Off,5
Tum-Off Current Gain, 12
Tum-Off Mechanism, 4
Tum-Off Methods, 127
Tum-Off Switch, 12
Tum-Off Time, 5, 10, 124
Tum-Off Time Extender,247-248
INDEX
Tum-OH Time Test, 589
Tum-OH Time Tester, 590
Turn-OH Time Variation, 125, 126
Tum-On, 2, 71,88,90, 74,492
Tum-On Time, Gate Controlled, 584
Tum-On Time Tester, 584
Turn-On Voltage, 55
Tum-On Voltage Test, 583
Turn-On Voltage Tester, 583
Two Transistor Analogy, 1
UJT,183
UJT, Programmable, 25
UJT, Temperature Coefficient, 101
UJT, Transient Suppression, 498
Unijunction Specifications, 634-636
Unijunction Transistors, 100, 183
Unit Pak Package Mounting, 529
Universal Motors, 287
UPS Inverter, 383
Varistor, Metal Oxide, 159
V(BO) Triggering, di/dt
Capability, 55
VOE,50l
VOM,63
VORM, OH-State BlockingVoltage, 61
Vehicle, Electric Controller, 369
Voltage, Blocking, 5,10
Voltage - Current Characteristics,
Triac, 183
Voltage, Forward Break-Over, 11,81
Voltage, Gate Trigger, Test, 570
Voltage, Measurement, 602
Voltage, Non-Repetitive, 29
Voltage, Peak OH-State, 567
Voltage, Peak On-Stilte, Test, 579
Voltage, Peak Point, 99
Voltage, Peak Reverse, 236-240, 567
Voltage Protective Circuits, 209, 212
Voltage Ratings, 60
Voltage Ratings, High Frequency, 63
Voltage Regulation, 355, 359, 383,
387,396
Voltage Regulator, 262
Voltage Sensing Circuit, 224
Voltage Sharing, 159
Voltage Transients, 469
Voltage, Tum-On, 55
Voltage, Tum-On, Test, 583
Voltage, Tum-On, Tester, 583
Voltmeter, Peak Reading, 569
Volume Requirements, Heat
Exchanger, 542
VRRM& VRSM (Reverse Voltage), 61
Washers, Belleville, 519
Water Cooling, 547
Water Properties for Liquid
Cooling, 549
Watt Second Loss Curves, 58,60
Wave Length, 423
Welding, Current Rating, 49
Wiring Layout, 490
X/R Ratio, 452
Zero-Voltage Switch, Light
Activated, 435
Zero-Voltage Switching, 201, 307,
334,497
Zero-Voltage Switching
Circuits, CSCR, 312
Zero Voltage Switching Circuits,
Discrete Transistor, 310, 312, 315
Zero Voltage Switching·
Circuits, IC, 316
Zero Voltage Switching
Circuits, SCR, 313
Zero Voltage SWitching,
High Frequency, 320
Zero Voltage Switching,
Three Phase, 321
687
NOTES
Source Exif Data:
File Type : PDF File Type Extension : pdf MIME Type : application/pdf PDF Version : 1.3 Linearized : No XMP Toolkit : Adobe XMP Core 4.2.1-c041 52.342996, 2008/05/07-21:37:19 Create Date : 2017:09:14 17:23:45-08:00 Modify Date : 2017:09:14 17:46:29-07:00 Metadata Date : 2017:09:14 17:46:29-07:00 Producer : Adobe Acrobat 9.0 Paper Capture Plug-in Format : application/pdf Document ID : uuid:09fd215a-3611-0843-bfe1-9e02c717a461 Instance ID : uuid:58d153d1-774d-4d4b-81ad-208404c07666 Page Layout : SinglePage Page Mode : UseNone Page Count : 712EXIF Metadata provided by EXIF.tools